GIGA CRM Phase 1 β€” PRD

INTUS β€” AI Opportunity Platform
ORG: INTUS Status: Draft v0.2 Revised: 2026-07-11 Charter: Approved 2026-06-23, Rev.1 2026-07-09

GIGA CRM Phase 1 β€” Product Requirements Document (PRD)

Organization: INTUS Document Status: Draft v0.2 (Revised based on charter original text) Created: 2026-07-10 Revised: 2026-07-11 Project Charter Approval: 2026-06-23 (Approved), 2026-06-29 (Date Approved) Charter Revision: Rev.1 β€” 2026-07-09 (Timeline shortened) Charter Reference: https://chatgpt.com/share/e/6a3a7106-6070-8000-ba0f-51d1f37b67c9 Document Owner: [TBD] Approver: [TBD]


Revision History

Version Date Author Changes
v0.1 2026-07-10 β€” Initial framework, based on project charter context reconstruction
v0.2 2026-07-11 β€” Comprehensive revision based on charter image verbatim reading: INTUS company name, product positioning correction, 9 quantifiable KPIs, 7 risks, 12 assumptions, 7 constraints, budget, stakeholders, architecture diagram, etc.

Table of Contents

  1. Project Overview
  2. Business Background & Problem Statement
  3. Project Goals & Success Criteria
  4. Scope Definition
  5. Stakeholders & User Roles
  6. Functional Requirements
  7. Non-Functional Requirements
  8. System Architecture Overview
  9. Data Requirements
  10. AI / Agent Requirements
  11. Integration Requirements
  12. Deliverables Mapping
  13. Milestones & Timeline
  14. Assumptions, Constraints & Risks
  15. Financial Information
  16. Acceptance Criteria
  17. SOX / Internal Control Impact
  18. Glossary
  19. Appendix

1. Project Overview

1.1 Project Name

GIGA CRM β€” AI Opportunity Platform (Phase 1)

1.2 Organization

INTUS

1.3 Project Positioning

GIGA CRM is part of a broader multi-phase roadmap to consolidate commercial workflows and build an agent-first CRM foundation. Phase 1 is limited to the AI Opportunity Platform, focused on agent-assisted and agent-operated workflows for Schedulers and Researchers.

Key Positioning Statement (from Charter Assumption #1): GIGA CRM is expected to replace or consolidate fragmented commercial systems over multiple phases, rather than add another standalone CRM, and will become the primary system for managing customer information and commercial activities.

Key Positioning Statement (from Charter Business Need #3): Phase 1 creates reusable foundations for future CRM phases, while not replacing the full CRM platform in Phase 1.

1.4 Document Purpose

This document is based on the approved Project Charter (Approved 2026-06-23, Rev.1 2026-07-09). It decomposes the Phase 1 scope defined in the charter into executable functional and non-functional requirements, serving as the basis for solution design, development, and acceptance testing.

1.5 Target Audience


2. Business Background & Problem Statement

2.1 Business Need (from Charter)

INTUS currently relies on disconnected systems and manual processes for opportunity research, scheduling, and commercial workflow coordination. This creates:

Pain Point Impact
Duplicated effort Same opportunity processed across multiple systems
Inconsistent information Data out of sync across systems, conflicting information
Limited visibility Management cannot view opportunity pipeline and team progress in real-time
Delays in preparing opportunities for Sales Extended cycle from opportunity discovery to Sales-ready handoff

2.2 Phase 1 Solution

Phase 1 will establish an AI Opportunity Platform for Researchers and Schedulers to improve: - Opportunity intake - Research - Scheduling coordination - Structured handoffs - Workflow visibility

2.3 Phase 1 Intent


3. Project Goals & Success Criteria

3.1 Phase 1 Goal

Establish an AI Opportunity Platform serving Researchers and Schedulers, implementing agent-assisted and agent-operated workflows, validating reusable workflow, data, AI governance, access control, and decision-traceability patterns.

3.2 Success Criteria (from Charter, 9 Quantifiable KPIs)

ID Success Dimension Target Measurement
SC-01 Research Productivity From 30 projects/5 days β†’ 50 projects/5 days Researchers generate structured opportunity research summaries using the platform
SC-02 Scheduling Efficiency 4 booked discovery meetings/day Schedulers manage meeting coordination, reminders, follow-up through the workflow
SC-03 Human Review Control β‰₯ 90% autonomous, ≀ 10% human intervention Items with unclear data, low confidence, scheduling conflicts, or priority flags route to human review; system processes β‰₯90% of pipeline decision points autonomously
SC-04 Data Consistency Template deviation ≀ 10% Research and scheduling outputs use standardized templates rather than freeform notes
SC-05 Workflow Visibility No manual report creation required Managers can view task volume, status, bottlenecks, cycle time, completion rate; access dashboard metrics directly
SC-06 Handoff Quality β‰₯ 95% complete handoff packages (at 50 projects/5 days) Sales or downstream teams receive complete handoff package with research context, scheduling status, and open questions
SC-07 Agent Accountability 100% of agent actions contain traceable decision logs Agent actions and recommendations include basic decision trace and confidence indicator
SC-08 AI Output Acceptance Rate β‰₯ 90% accepted without major rework AI-generated research outputs accepted without major rework during pilot
SC-09 Pilot Adoption 100% of designated users actively using weekly Initial Scheduler and Researcher users actively use the platform during Phase 1 pilot

4. Scope Definition

4.1 Phase 1 In-Scope (from Charter)

4.2 Phase 1 Out-of-Scope (from Charter)

Excluded Item Note
Full CRM replacement β€”
Quote management β€”
Opportunity pipeline management β€”
Customer service workflows β€”
ERP replacement β€”
External customer-facing AI communication β€”
Autonomous sales negotiation β€”
L0 Reviewer functionality Phase II
Deal Coach functionality Phase III
Full commercial platform rollout β€”
Revenue forecasting automation β€”
Other department workflows β€”

4.3 Data Source Scope

4.4 Future Roadmap Note

From Charter: Future CRM roadmap items are directional only and NOT approved deliverables under this Phase 1 charter. Future phases require separate scope approval, prioritization, timeline confirmation, and funding review.

4.5 Scope Boundary Diagram

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚                   GIGA CRM Phase 1 Scope                       β”‚
β”‚              AI Opportunity Platform                          β”‚
β”‚                                                              β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”    β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”    β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”           β”‚
β”‚  β”‚ Data      │───▢│ Researcher │───▢│ Scheduler β”‚           β”‚
β”‚  β”‚ Intake    β”‚    β”‚ Workflow   β”‚    β”‚ Workflow  β”‚           β”‚
β”‚  β”‚ Dodge/CC  β”‚    β”‚ AI-assistedβ”‚    β”‚ AI-advisedβ”‚           β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜    β””β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”˜    β””β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”˜           β”‚
β”‚                         β”‚                β”‚                    β”‚
β”‚                         β–Ό                β–Ό                    β”‚
β”‚                  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”             β”‚
β”‚                  β”‚ Human      β”‚  β”‚ Structured  β”‚             β”‚
β”‚                  β”‚ Review Gateβ”‚  β”‚ Handoff Pkg β”‚             β”‚
β”‚                  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜  β””β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”˜             β”‚
β”‚                                        β”‚                     β”‚
β”‚                         β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β–Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”      β”‚
β”‚                         β”‚ Status Tracking + Dashboard  β”‚      β”‚
β”‚                         β”‚ + Notifications              β”‚      β”‚
β”‚                         β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜      β”‚
β”‚                                                              β”‚
β”‚  ──────────── Phase 1 Boundary ────────────                  β”‚
β”‚  βœ… Researchers + Schedulers workflows                        β”‚
β”‚  βœ… Agent-assisted + Agent-operated modes                     β”‚
β”‚  ❌ Full CRM / Quotes / Pipeline / CS / ERP                  β”‚
β”‚  ❌ External customer-facing AI / Autonomous sales            β”‚
β”‚  ❌ L0 Reviewer (Phase II) / Deal Coach (Phase III)         β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
     Data Sources: Dodge | ConstructConnect (only)

5. Stakeholders & User Roles

5.1 Project Stakeholders (from Charter)

Role Name Responsibility
CEO (Approval Committee) Roland Talalas Final project approval
PMO (Approval Committee) Kimia Soroush Project management governance
Product Owner (Approval Committee) Kurtis Perdelwitz Product requirement decisions
Project Stakeholder Ryan Gombar Business participation
Project Stakeholder Gabriele Daugmaudyte Business participation
Project Stakeholder Connor McGorty Business participation
Project Stakeholder GiedrΔ— JankauskienΔ— Business participation

5.2 User Role Definitions (from Charter)

Researcher

Users responsible for opportunity background research

Attribute Description
Core Responsibilities Review AI-filtered opportunities; enrich and evaluate opportunities; review/edit AI research summaries; determine opportunity priority; adjust AI instructions or criteria and re-trigger AI processing
Outputs Standardized research summaries, evaluation results, confidence indicators, recommended next actions
Key Permissions Can adjust AI processing instructions/criteria and re-trigger; can edit AI-generated summaries
Pain Points Manual screening time-consuming; information scattered across systems; lack of unified priority standards

Scheduler

Users responsible for meeting coordination, reminders, follow-up, and related scheduling workflows

Attribute Description
Core Responsibilities Review and adjust AI-prepared communication plans; manage email drafts, voice call plans, timing, follow-up sequence, escalation notes; execute Human Review Gate reviews
Outputs Reviewed communication plans, executed communication records, handoff confirmations
Key Permissions Can adjust AI-generated communication plans; approve/reject customer-facing communication at Human Review Gate
Pain Points Incomplete handoff information; scheduling relies on experience without data support; status tracking via spreadsheets with poor real-time visibility

Manager

Attribute Description
Core Responsibilities Monitor team performance, view dashboards, approve key decisions
Key Permissions Access dashboard metrics without manual report creation; view task volume, status, bottlenecks, cycle time, completion rate

Admin

Attribute Description
Core Responsibilities System configuration, user management, permission management, data source configuration

5.3 Role Access Control Matrix

Functional Area Researcher Scheduler Manager Admin
Opportunity list view βœ… βœ… βœ… βœ…
AI filtering configuration ❌ ❌ View βœ…
Opportunity enrichment & evaluation βœ… View View βœ…
AI summary editing βœ… ❌ View βœ…
Communication plan review View βœ… View βœ…
Human Review Gate ❌ βœ… βœ… βœ…
Handoff package confirmation ❌ βœ… View βœ…
Dashboard access Limited Limited βœ… βœ…
Audit log view ❌ ❌ βœ… βœ…

6. Functional Requirements

6.1 Module Overview

Module ID Module Name Deliverable Priority
F-01 AI Opportunity Platform Foundation D-01 P0
F-02 Data Ingestion D-02 P0
F-03 Opportunity Filtering D-03 P0
F-04 Opportunity Enrichment & Evaluation Review D-04 P0
F-05 Communication Plan Review D-05 P1
F-06 Human Review Gate D-06 P0
F-07 Structured Handoff Package D-07 P1
F-08 Reporting & Visibility D-08 P2
F-09 Pilot Enablement D-09 P1
F-10 Workflow Status Tracking D-10 P0
F-11 Notifications & Reminders D-11 P1

6.2 F-01: AI Opportunity Platform Foundation

Goal: Establish initial platform workspace for managing opportunity-related research and scheduling workflows with basic access controls.

ID Requirement Priority Acceptance Criteria
F-01-01 Platform workspace initialization P0 Initial platform workspace for managing opportunity-related research and scheduling workflows
F-01-02 Role-based access control P0 Basic access controls for Scheduler, Researcher, Manager, and Admin roles
F-01-03 Role permission mutual exclusion P0 Different roles have mutually exclusive permissions; users only see functions and data allowed by their role
F-01-04 Audit log recording P0 All key operations have audit log records, queryable, tamper-proof
F-01-05 Session management P0 Session timeout (30 minutes idle auto-logout)
F-01-06 System configuration management P1 Admin can configure data source connections, AI parameters, filtering rules, etc.

6.3 F-02: Data Ingestion

Goal: Import Dodge and ConstructConnect project data, normalize into a unified structure, deduplicate records, and store for downstream processing.

ID Requirement Priority Acceptance Criteria
F-02-01 Dodge data import P0 Support automatic Dodge project data import, daily scheduled + manual trigger
F-02-02 ConstructConnect data import P0 Support automatic ConstructConnect project data import, daily scheduled + manual trigger
F-02-03 Data normalization P0 External fields auto-mapped to internal unified data structure
F-02-04 Data deduplication P0 Same project from multiple sources auto-identified and merged
F-02-05 Data storage P0 Normalized data stored for downstream processing
F-02-06 Import exception handling & alerting P0 Import failures notify admin, support retry, log exceptions
F-02-07 Import history records P1 View each import's records, quantity, status, timestamp

Data Ingestion Flow:

Dodge API ──┐
             β”œβ”€β”€β–Ά Field Mapping/Normalize ──▢ Deduplicate ──▢ Store ──▢ Downstream
ConstructConnect API β”€β”€β”˜                          β”‚
                                                   β”œβ”€β”€ Error? ──▢ Alert Admin
                                                   └── Success? ──▢ Log Record

6.4 F-03: Opportunity Filtering

Goal: Apply agreed business criteria to filter opportunities before AI enrichment and evaluation.

ID Requirement Priority Acceptance Criteria
F-03-01 Business filtering criteria configuration P0 Admin can configure filtering conditions (industry, region, scale, stage, etc.)
F-03-02 Automatic filtering execution P0 Apply filtering criteria before AI enrichment and evaluation
F-03-03 Filtered results list view P0 Support list/card view, sortable, filterable, paginated
F-03-04 Filtering rule version management P1 Rule changes traceable, support rollback
F-03-05 Batch operations P1 Support batch accept/reject/priority adjustment
F-03-06 Filtered results export P2 Support Excel/CSV export

6.5 F-04: Opportunity Enrichment & Evaluation Review

Goal: Review and correction loop for AI-prepared opportunity enrichment, qualification context, product fit, confidence indicators, and recommended next actions. Researchers can adjust instructions or criteria and re-trigger AI processing.

ID Requirement Priority Acceptance Criteria
F-04-01 AI automatic information enrichment P0 AI auto-supplements contacts, company info, historical records
F-04-02 AI qualification context P0 Generate qualification context for each opportunity
F-04-03 AI product fit assessment P0 Evaluate opportunity-product fit
F-04-04 AI confidence indicators P0 Each evaluation result includes confidence indicators
F-04-05 AI recommended next actions P0 Generate recommended next actions
F-04-06 AI research summary generation P0 One-click structured opportunity research summary
F-04-07 Researcher review/correction loop P0 Researcher can review and correct AI results
F-04-08 Re-trigger AI processing P0 Researcher can adjust instructions or criteria and re-trigger AI processing
F-04-09 AI summary manual editing P0 Researcher can edit/correct AI-generated summaries
F-04-10 Evaluation history P1 Each evaluation result traceable, support version comparison
F-04-11 Priority label management P1 Support High/Medium/Low and custom labels

6.6 F-05: Communication Plan Review

Goal: Workflow for Schedulers to review and adjust AI-prepared communication plans, including email drafts, voice call plans, timing, follow-up sequence, and escalation notes.

ID Requirement Priority Acceptance Criteria
F-05-01 AI communication plan generation P1 Generate suggested communication plan based on opportunity info (email drafts, voice plans, timing, follow-up sequence, escalation notes)
F-05-02 Email draft review & adjustment P1 Scheduler can review and edit AI-generated email drafts
F-05-03 Voice call plan review P1 Scheduler can review and adjust voice call plans
F-05-04 Timing review P1 Scheduler can review and adjust communication timing
F-05-05 Follow-up sequence review P1 Scheduler can review and adjust follow-up sequence
F-05-06 Escalation notes review P1 Scheduler can review and adjust escalation notes
F-05-07 Communication plan template management P2 Pre-built templates, support customization
F-05-08 Communication record tracking P1 Record each communication result, linked to opportunity

6.7 F-06: Human Review Gate

Goal: Required review step before customer-facing communication is sent or initiated.

ID Requirement Priority Acceptance Criteria
F-06-01 Required review for customer-facing communication P0 Customer-facing communication must be reviewed before sending or initiating
F-06-02 Review gate configuration P0 Admin can configure which nodes require human review
F-06-03 Review task assignment P0 Review tasks auto-assigned to appropriate roles
F-06-04 Review interface P0 Display AI decision context, confidence indicators, source visibility; support approve/reject/return
F-06-05 Exception management P0 Unclear data, low confidence, scheduling conflicts, or priority flags route to human review
F-06-06 Review SLA management P1 Set review time limits, auto-escalate on timeout
F-06-07 Review feedback loop P1 Review results fed back to AI model for continuous optimization

Review Gate Nodes:

AI Enrichment/Evaluation ──▢ [Gate] Researcher Review (can adjust/re-trigger)
                                        β”‚
AI Communication Plan ────▢ [Gate] Scheduler Review (email/voice/timing/follow-up/escalation)
                                        β”‚
Customer-facing Comm ────▢ [Gate] Human Review Gate (required before sending)
                                        β”‚
Handoff Package ─────────▢ [Gate] Confirm handoff package completeness

6.8 F-07: Structured Handoff Package

Goal: Track approved email and voice communications, generate structured handoff packages for Sales.

ID Requirement Priority Acceptance Criteria
F-07-01 Handoff package auto-generation P1 AI aggregates opportunity info to generate handoff package
F-07-02 Standardized handoff package template P0 Includes project context, enrichment, qualification result, outreach activity, response status, scheduling status, open questions
F-07-03 Handoff confirmation flow P0 Send β†’ Recipient confirms β†’ Complete/Return
F-07-04 Handoff package version management P1 Support multiple versions, return and modify generates new version
F-07-05 Handoff statistics report P2 Statistics on handoff quantity, return rate, average processing time

Handoff Package Structure (from Charter):

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚       Structured Handoff Package          β”‚
β”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€
β”‚ 1. Project Context                        β”‚
β”‚    - Project name, address, type           β”‚
β”‚    - Owner/GC information                  β”‚
β”‚                                          β”‚
β”‚ 2. Enrichment                             β”‚
β”‚    - AI enrichment + Researcher edits      β”‚
β”‚    - Contacts, company information        β”‚
β”‚                                          β”‚
β”‚ 3. Qualification Result                   β”‚
β”‚    - Qualification context                β”‚
β”‚    - Product fit                          β”‚
β”‚    - Confidence indicators                β”‚
β”‚    - Priority                             β”‚
β”‚                                          β”‚
β”‚ 4. Outreach Activity                      β”‚
β”‚    - Approved email/voice communication   β”‚
β”‚    - Communication plan & execution statusβ”‚
β”‚                                          β”‚
β”‚ 5. Response Status                        β”‚
β”‚    - Customer response records            β”‚
β”‚    - Follow-up status                     β”‚
β”‚                                          β”‚
β”‚ 6. Scheduling Status                      β”‚
β”‚    - Meeting arrangement status           β”‚
β”‚    - Follow-up sequence progress          β”‚
β”‚                                          β”‚
β”‚ 7. Open Questions                         β”‚
β”‚    - Items to be confirmed                β”‚
β”‚    - Risk alerts                          β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜

6.9 F-08: Reporting & Visibility

Goal: Initial dashboard showing opportunity volume, workflow status, cycle time, blocked items, completed tasks, and human review rate.

ID Requirement Priority Acceptance Criteria
F-08-01 Opportunity volume dashboard P1 Display total opportunities, distribution by stage
F-08-02 Workflow status view P1 Display opportunity count by workflow status
F-08-03 Cycle time analysis P1 Display average cycle time from import to handoff
F-08-04 Blocked items view P1 Display currently blocked opportunities and reasons
F-08-05 Completed tasks statistics P1 Display completed task count and trends
F-08-06 Human review rate P1 Display human review proportion, review pass rate
F-08-07 No manual reports required P0 Managers can access dashboard metrics directly without manual report creation
F-08-08 Report export P2 Support PDF/Excel export

6.10 F-09: Pilot Enablement

Goal: Training and feedback loop for initial Researcher and Scheduler users.

ID Requirement Priority Acceptance Criteria
F-09-01 User training P0 Provide training for initial Researcher and Scheduler users
F-09-02 Feedback loop P0 Establish user feedback collection and processing mechanism
F-09-03 User manual P1 Operation guide covering all functional modules
F-09-04 In-app onboarding hints P2 Guided tutorial for new users' first use
F-09-05 Pilot support channel P0 Provide ticket/instant messaging support channel

6.11 F-10: Workflow Status Tracking

Goal: Track research and scheduling work statuses with audit traceability.

Workflow Statuses (from Charter):

Status Description
New Opportunity just imported, not yet processed
In Progress Currently being processed
Needs Human Review Human review condition triggered, awaiting review
Ready for Handoff Processing complete, handoff package can be generated
Completed Handoff complete, workflow ended
Blocked Blocked for some reason, requires intervention
ID Requirement Priority Acceptance Criteria
F-10-01 State machine definition P0 Clearly define 6 statuses and transition rules
F-10-02 Automatic status transition P0 AI Agent advances status based on workflow
F-10-03 Manual status intervention P0 Users can manually adjust status (with permissions)
F-10-04 Status change history P0 Complete status change records (who, when, why)
F-10-05 Status board view P1 Kanban-style display of opportunities by status
F-10-06 Timeout alert P1 Alert when opportunity stays in a status beyond threshold

6.12 F-11: Notifications & Reminders

Goal: Basic notifications for overdue tasks, pending reviews, scheduling conflicts, and required human approvals.

ID Requirement Priority Acceptance Criteria
F-11-01 Overdue task notification P0 Notify responsible person when task is overdue
F-11-02 Pending review notification P0 Notify relevant personnel when review tasks are pending
F-11-03 Scheduling conflict notification P0 Notify Scheduler when scheduling conflict detected
F-11-04 Required human approval notification P0 Notify approver when human approval is required
F-11-05 Notification preferences P2 Users can configure delivery method and frequency
F-11-06 Daily digest email P2 Send daily work summary to users

7. Non-Functional Requirements

7.1 Performance

ID Requirement Metric
NFR-P-01 Page load time First screen ≀ 3s (P95)
NFR-P-02 API response time Normal ≀ 500ms (P95), complex query ≀ 2s (P95)
NFR-P-03 AI summary generation time ≀ 30s (P95)
NFR-P-04 Data import throughput Single import β‰₯ 5,000 opportunities
NFR-P-05 Concurrent user support β‰₯ 100 concurrent users

7.2 Availability

ID Requirement Metric
NFR-A-01 System availability β‰₯ 99.5% (monthly)
NFR-A-02 Planned maintenance window Monthly, 48-hour advance notice
NFR-A-03 Recovery Time Objective (RTO) ≀ 4 hours
NFR-A-04 Recovery Point Objective (RPO) ≀ 1 hour

7.3 Security

ID Requirement Description
NFR-S-01 Data transmission encryption HTTPS / TLS 1.2+
NFR-S-02 Data storage encryption Sensitive fields encrypted at rest
NFR-S-03 Authentication Support SSO (OAuth 2.0 / SAML)
NFR-S-04 Least privilege Role-based fine-grained access control (RBAC)
NFR-S-05 Audit logs Retained β‰₯ 1 year, tamper-proof
NFR-S-06 Periodic access reviews Regular access permission reviews

7.4 Scalability

ID Requirement Description
NFR-E-01 Horizontal scaling Core services support horizontal scaling, stateless design
NFR-E-02 Data source extensibility Data ingestion layer supports plugin-based new data source addition (requires formal approval)
NFR-E-03 AI model extensibility AI inference layer supports model hot-swap
NFR-E-04 Reusable patterns Phase 1 architecture should create reusable workflow, data, and AI governance patterns

7.5 Compatibility

ID Requirement Description
NFR-C-01 Browser compatibility Chrome 100+, Edge 100+, Safari 15+
NFR-C-02 Resolution Min 1280Γ—720, recommended 1920Γ—1080
NFR-C-03 Responsive design Support tablet landscape (basic adaptation)

8. System Architecture Overview

8.1 Architecture Overview (from Charter)

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚                        User Layer                                β”‚
β”‚    Researchers  |  Schedulers  |  Inside Sales Rep  |  PMs      β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
                            β”‚
β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β–Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚                   AI-first CRM Platform                          β”‚
β”‚                                                                  β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚              AI Agent Runtime                               β”‚ β”‚
β”‚  β”‚        (Isolated Docker Sandboxes)                          β”‚ β”‚
β”‚  β”‚   β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”        β”‚ β”‚
β”‚  β”‚   β”‚ Data    β”‚ β”‚ Eval    β”‚ β”‚ Corres  β”‚ β”‚ Researchβ”‚  ...    β”‚ β”‚
β”‚  β”‚   β”‚ Ingest  β”‚ β”‚ Agent   β”‚ β”‚ pondenceβ”‚ β”‚ Agent   β”‚        β”‚ β”‚
β”‚  β”‚   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜        β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                             β”‚                                    β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β–Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚          CRM Core Operational Database                      β”‚ β”‚
β”‚  β”‚             (System of Record)                              β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                             β”‚                                    β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β–Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚              Operational Modules                            β”‚ β”‚
β”‚  β”‚   β”Œβ”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β”          β”‚ β”‚
β”‚  β”‚   β”‚Importβ”‚ β”‚Filterβ”‚ β”‚Enrichβ”‚ β”‚Handoffβ”‚ β”‚Reportβ”‚          β”‚ β”‚
β”‚  β”‚   β””β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”˜ β””β”€β”€β”€β”€β”€β”€β”˜          β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                             β”‚                                    β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β–Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚                    Data Layer                               β”‚ β”‚
β”‚  β”‚   Low Level Adapters | Data Warehousing | Analytics         β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                                                                  β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚              Data Governance Layer                          β”‚ β”‚
β”‚  β”‚   Users | Accounts | Permissions | Standard Ingestion       β”‚ β”‚
β”‚  β”‚   Common API | Middleware | Analytics                       β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                                                                  β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚
β”‚  β”‚              Agent Governance                               β”‚ β”‚
β”‚  β”‚      Agent lifecycle, behavior, and safety                  β”‚ β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚
β”‚                                                                  β”‚
β”‚  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”  β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”  β”‚
β”‚  β”‚ Phase 1 Scope   β”‚  β”‚ Future Phases (Not Yet in Phase)    β”‚  β”‚
β”‚  β”‚ (This Phase)    β”‚  β”‚ L0 Reviewer, Deal Coach, etc.       β”‚  β”‚
β”‚  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜  β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜  β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜

8.2 Architecture Layer Description

Layer Name Responsibility Key Feature
β€” User Layer Researchers, Schedulers, Inside Sales Rep, PMs/Delivery Multi-role access
L1 AI Agent Runtime AI Agent execution environment Docker sandbox isolation; Phase 1 Agents
L2 CRM Core Operational Database Core business data storage System of Record
L3 Operational Modules Business functional modules Import, filter, enrich, handoff, report, etc.
L4 Data Layer External data access and storage Low Level Adapters, Data Warehousing, Analytics & Insights
β€” Data Governance Layer Data governance Users, Accounts, Permissions, Standard Ingestion, Common API, Middleware, Analytics
β€” Agent Governance Agent governance Agent lifecycle, behavior, and safety

8.3 Architecture Principles

8.4 Key Architecture Decisions (TBD in Solution Design)

Decision Option A Option B Recommendation
Deployment architecture Modular monolith Microservices Phase 1: modular monolith to reduce complexity
AI inference Cloud API On-premise Hybrid: summaries via cloud API, filtering via local model
Frontend framework React Vue Based on team tech stack
Message queue RabbitMQ Kafka Phase 1: RabbitMQ, sufficient

9. Data Requirements

9.1 Core Data Entities

Entity Description Key Fields Source
Opportunity Core business object ID, project name, address, type, stage, score, status Dodge / CC
Contact Opportunity-related contact ID, name, role, phone, email, company Enrichment / Manual
Company Owner/GC company ID, name, address, type, industry Enrichment / Manual
Enrichment AI enrichment result ID, opportunity ID, enrichment content, confidence, source AI Agent
Assessment Opportunity assessment record ID, opportunity ID, assessor, qualification, product fit, confidence, recommended actions AI + Human
CommunicationPlan AI-prepared communication plan ID, opportunity ID, email drafts, voice plans, timing, follow-up sequence, escalation notes, review status AI + Scheduler
HandoffPackage Structured handoff package ID, opportunity ID, project context, enrichment, qualification, outreach activity, response status, scheduling status, open questions System
AgentTask AI Agent executed task ID, type, input, output, status, decision trace, confidence, timestamp System
AuditLog Operation audit ID, user, operation, object, timestamp, details System

9.2 Data Retention Policy

Data Type Retention Archive Strategy
Active opportunity data Permanent β€”
Closed opportunities 3 years Archive to cold storage after 3 years
Audit logs 1 year Compress and archive after 1 year
Agent task records 6 months Aggregate statistics after 6 months, delete details
AI summary history 2 years Keep only final version after 2 years
Decision trace logs β‰₯ 1 year Tamper-proof, used for traceability

10. AI / Agent Requirements

10.1 AI Agent Definitions

Agent Name Responsibility Input Output Mode
Data Ingestion Agent Data import & normalization Dodge/CC raw data Normalized unified structure data Agent-operated
Filtering Agent Opportunity filtering Opportunity pool + business criteria Filtered opportunity list Agent-operated
Enrichment Agent Information enrichment Opportunity basic info Supplemented contacts/company info + confidence Agent-assisted
Evaluation Agent Qualification assessment Opportunity + enrichment Qualification context + product fit + confidence + recommended actions Agent-assisted
Research Summary Agent Research summary generation Opportunity + enrichment + assessment Structured research summary Agent-assisted
Communication Plan Agent Communication plan generation Opportunity + assessment Email drafts + voice plans + timing + follow-up sequence + escalation notes Agent-assisted
Scheduling Agent Scheduling recommendations Opportunity + history + rules Scheduling recommendation plan Agent-assisted

10.2 Agent Operating Modes (from Charter)

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚         Agent-assisted Mode (Phase 1 Priority)   β”‚
β”‚  AI operates within working boundaries            β”‚
β”‚  AI provides suggestions/drafts β†’ Human review    β”‚
β”‚  External communication requires approval         β”‚
β”‚  Humans handle exceptions, approvals, high-value  β”‚
β”‚  decisions                                        β”‚
β”‚  Applicable: Enrichment, evaluation, summaries,   β”‚
β”‚  communication plans, scheduling recommendations   β”‚
β”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€
β”‚         Agent-operated Mode                       β”‚
β”‚  AI executes autonomously β†’ Notifies human        β”‚
β”‚  Applicable: Data import, filtering, enrichment,  β”‚
β”‚  status transitions                               β”‚
β”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€
β”‚         Human-gated Mode                          β”‚
β”‚  AI prepares β†’ Human approval required to proceed β”‚
β”‚  Customer-facing communication must be reviewed   β”‚
β”‚  before sending                                   β”‚
β”‚  Applicable: Handoff package, customer-facing com β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜

10.3 Agent Governance (from Charter)

10.4 Agent Authority Expansion Conditions (from Charter)

Agent authority will not expand unless quality, traceability, and review thresholds are documented and met.

10.5 AI Quality Requirements

Metric Target Measurement
AI output acceptance rate β‰₯ 90% AI-generated research outputs accepted without major rework during pilot
Confidence indicator coverage 100% All Agent actions include confidence indicators
Source visibility 100% All AI outputs include source visibility
Decision trace coverage 100% All Agent actions include traceable decision logs
Human intervention rate ≀ 10% Human intervention does not exceed 10% of total processed
AI response time ≀ 30s P95 latency

10.6 AI Feedback Loop

AI Output ──▢ Human Review ──▢ Review Result
                                 β”‚
                          β”Œβ”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”
                          β”‚ Accept       β”‚ Reject/Modify
                          β”‚              β”‚
                          β–Ό              β–Ό
                    Record positive  Record negative + correction
                          β”‚              β”‚
                          β””β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”˜
                                 β”‚
                          Periodic model retraining / Prompt optimization

11. Integration Requirements

11.1 External System Integrations

System Integration Method Data Flow Frequency Priority
Dodge REST API Pull opportunity data Daily P0
ConstructConnect REST API Pull opportunity data Daily P0
Email System (SMTP) SMTP Send notifications/email drafts Event-triggered P0
Voice Communication Platform API/Integration Voice call plan execution Event-triggered P1
SSO OAuth 2.0 / SAML User authentication Real-time P0
File Storage SDK Store attachments Real-time P1

11.2 Integration Constraints

From Charter: Integration feasibility and pilot readiness depend on source system capability, API maturity, and data quality.

From Charter: Existing source systems are expected to contain sufficient data quality for meaningful AI-assisted research outputs.

11.3 Integration Interface Specification

Dodge API Integration
β”œβ”€β”€ Endpoint: [TBD]
β”œβ”€β”€ Authentication: API Key
β”œβ”€β”€ Frequency: Daily 02:00 UTC
β”œβ”€β”€ Data Format: JSON
β”œβ”€β”€ Error Handling: Retry 3 times, 5-minute interval
└── Alerting: Notify admin on 2 consecutive failures

ConstructConnect API Integration
β”œβ”€β”€ Endpoint: [TBD]
β”œβ”€β”€ Authentication: API Key
β”œβ”€β”€ Frequency: Daily 03:00 UTC
β”œβ”€β”€ Data Format: JSON
β”œβ”€β”€ Error Handling: Retry 3 times, 5-minute interval
└── Alerting: Notify admin on 2 consecutive failures

12. Deliverables Mapping

12.1 Charter Deliverables to PRD Requirements Mapping (from Charter)

Deliverable ID Deliverable Name Charter Description Module Requirements
D-01 AI Opportunity Platform Foundation Initial platform workspace for managing opportunity-related research and scheduling workflows. Basic access controls for Scheduler, Researcher, manager, and admin roles F-01 F-01-01 ~ F-01-06
D-02 Data Ingestion Import Dodge and ConstructConnect project data, normalize into unified structure, deduplicate, store for downstream processing F-02 F-02-01 ~ F-02-07
D-03 Opportunity Filtering Apply agreed business criteria to filter opportunities before AI enrichment and evaluation F-03 F-03-01 ~ F-03-06
D-04 Opportunity Enrichment & Evaluation Review Review and correction loop for AI-prepared enrichment, qualification context, product fit, confidence indicators, recommended next actions. Researchers can adjust instructions or criteria and re-trigger AI processing F-04 F-04-01 ~ F-04-11
D-05 Communication Plan Review Workflow for Schedulers to review and adjust AI-prepared communication plans, including email drafts, voice call plans, timing, follow-up sequence, escalation notes F-05 F-05-01 ~ F-05-08
D-06 Human Review Gate Required review step before customer-facing communication is sent or initiated F-06 F-06-01 ~ F-06-07
D-07 Structured Handoff Package Tracking of approved email and voice communications, including project context, enrichment, qualification, outreach activity, response status, scheduling status, open questions F-07 F-07-01 ~ F-07-05
D-08 Reporting & Visibility Initial dashboard: opportunity volume, workflow status, cycle time, blocked items, completed tasks, human review rate F-08 F-08-01 ~ F-08-08
D-09 Pilot Enablement Training and feedback loop for initial Researcher and Scheduler users F-09 F-09-01 ~ F-09-05
D-10 Workflow Status Tracking Research and scheduling work statuses: New, In Progress, Needs Human Review, Ready for Handoff, Completed, Blocked F-10 F-10-01 ~ F-10-06
D-11 Notifications & Reminders Basic notifications for overdue tasks, pending reviews, scheduling conflicts, required human approvals F-11 F-11-01 ~ F-11-06

13. Milestones & Timeline

13.1 Milestone Overview (from Charter)

Milestone Target Key Activities Exit Criteria
M1: Solution Design Approved Jun 2026 Confirm scope, Dodge and ConstructConnect data intake, workflows, roles, success criteria Solution design approved
M2: Core Platform Operational Jul 2026 Platform setup, data intake, data model, access, security, workflow statuses Core platform operational
M3: AI Workflows Functional Aug 2026 Enrichment, evaluation, campaign planning, execution tracking, human review, handoff AI workflows functional
M4: UAT Complete / Pilot Readiness Sep 2026 End-to-end testing, user validation, defect fixes, pilot readiness review Go-live approval
M5: Pilot Go-Live Sep 2026 Pilot deployment for Researchers and Schedulers Pilot deployed
M6: Pilot Stabilized Oct 2026 Adoption monitored, feedback collected, priority issues resolved Pilot operational and stabilized
M7: Phase 1 Closure Oct 2026 Monitoring, feedback, issue resolution, KPI review, lessons learned, documentation, handover Phase 1 closed

⚠️ Note: Charter Rev.1 (2026-07-09) has shortened the timeline. Specific shortened dates need to be confirmed with the latest charter version.

13.2 Phase 1 Module-Milestone Mapping

Jul (M2)          Aug (M3)           Sep (M4-M5)        Oct (M6-M7)
─────────────────────────────────────────────────────────────────────
F-01 Platform    β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-02 Data Ingest β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-10 Status Trk  β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-03 Filtering             β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-04 Enrich/Eval           β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-05 Comm Plan             β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-06 Human Gate            β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-07 Handoff Pkg           β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-08 Reporting                       β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-09 Pilot Enable                    β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
F-11 Notifications                   β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ
                                     UAT        Pilot     Stabilize

13.3 Dependencies (from Charter)

The project depends on: - Dodge and ConstructConnect data being available - Required system connections working - Right user access setup - AI outputs being tested before launch - Researchers and Schedulers completing testing and training


14. Assumptions, Constraints & Risks

14.1 Assumptions (from Charter, 12 items)

ID Assumption
A-01 GIGA CRM is expected to replace or consolidate fragmented commercial systems over multiple phases, rather than add another standalone CRM, and will become the primary system for managing customer information and commercial activities
A-02 Phase 1 approval does not constitute approval for future CRM phases, broader commercial automation, revenue forecasting, quote management, ERP integration, or customer master data changes
A-03 Phase 1 will focus on internal users and Researcher and Scheduler workflows only
A-04 Phase 1 will use an agent-assisted model first, where AI operates within working boundaries (e.g., external communication requires approval) with humans handling exceptions, approvals, and high-value decisions
A-05 The pilot will use real users, real data, and real deals
A-06 Agent authority will not expand unless quality, traceability, and review thresholds are documented and met
A-07 Phase 1 should create reusable workflow, data, and AI governance patterns for future CRM phases
A-08 Future phases may expand into sales, quoting, opportunity management, service, forecasting, and broader commercial workflows
A-09 Future phases require separate scope approval, prioritization, timeline confirmation, and funding review
A-10 The Customer Service ERP will remain in place and integrate with CRM rather than be replaced
A-11 Researchers, Schedulers, Subject Matter Experts, and required support teams will be available for pilot validation and feedback
A-12 Existing source systems are expected to contain sufficient data quality for meaningful AI-assisted research outputs

14.2 Constraints (from Charter, 7 items)

ID Constraint
C-01 The Phase 1 timeline runs from Jun 2026 – Nov/Dec 2026
C-02 Integration feasibility and pilot readiness depend on source system capability, API maturity, and data quality
C-03 Agent-first operation requires defined governance, decision traceability, reversibility, and incident response before broad rollout
C-04 Customer-facing agent communication requires separate policy approval before use in pilot
C-05 Phase 1 must remain limited to the Researcher and Scheduler platform unless formal scope approval is obtained
C-06 Future CRM roadmap items remain directional and require separate approval before execution
C-07 Any change that expands Phase 1 into customer-facing autonomous communication, quoting, revenue forecasting, ERP integration, customer master data, financial reporting, or automated commercial commitments requires separate control-impact review before execution

14.3 Risk Register (from Charter, 7 items)

ID Risk Risk Owner Likelihood Impact Mitigation
R-01 Pipeline hygiene erodes after go-live Product Owner Medium High Launch live hygiene dashboard from day one; require weekly manager review and quarterly pruning
R-02 Scope creep from stakeholder requests PM and PO High High Publish out-of-scope list; require written trade-off for any addition
R-03 Researcher adoption fails despite on-time delivery Product Owner Medium High Start change management in month one; use pilot pod feedback as formal input to iteration
R-04 Required source systems do not provide sufficient data quality or integration capability PO Medium Medium Validate data availability, API capabilities, and integration requirements during discovery before committing to automation scope
R-05 Parallel run of old and new systems is painful Technical Lead High Medium Plan explicit cutover by module; pilot pod runs new only
R-06 AI-generated research contains inaccurate or outdated information Product Owner Medium High Require confidence scoring, source visibility, and human review triggers before handoff
R-07 Unauthorized access to opportunity data Technical Lead Low High Implement role-based permissions, audit logging, and periodic access reviews

15. Financial Information (from Charter)

15.1 Financial Assumptions

15.2 Cost Estimate

Cost Category Type Quantity/Assumption Estimated Cost
Soft costs Internal resources - AI/Product Engineering 3-4 person-months $12,000 - $16,000
Soft costs Internal resources - Product Analyst 2-3 person-months $6,000 - $9,000
Soft costs Internal resources - PMO Project Manager 3-4 person-months $6,000 - $9,000
Soft costs Internal resources - Researcher SME 0.25-0.5 person-months $1,000 - $3,000
Soft costs Internal resources - Scheduler SME 0.25-0.5 person-months $1,000 - $3,000
Hard costs External resources - AI/API consumption, email and voice communication platforms, integration tooling, hosting, monitoring Per month $100 - $300
Total $26,600 - $41,800

15.3 Budget Approval

Budget approved by Director of Finance.


16. Acceptance Criteria

16.1 Functional Acceptance Criteria

Acceptance criteria for each functional requirement are defined in Section 6. Overall functional acceptance requires:

16.2 UAT Acceptance Criteria

Acceptance Item Standard
Test case coverage β‰₯ 90% of functional requirements have test cases
Test pass rate β‰₯ 95%
Critical defects 0 P0 defects
General defects ≀ 5 P1 defects (with fix plan)
Pilot user acceptance Pilot users sign off

16.3 Go-Live Checklist


17. SOX / Internal Control Impact (from Charter)

17.1 Preliminary Classification

17.2 Control-Relevant Elements

Although no direct SOX impact, the project includes control-relevant elements:

Control Element Description
Role-based access RBAC permission control
Audit logging Operation audit records
Decision traceability Agent decision trace logs
Human review gates Required review for customer-facing communication
Customer-facing communication Must be reviewed before sending

17.3 Reassessment Conditions

If the project later expands to the following scope, SOX/control-impact classification must be reassessed: - Quote management - Revenue forecasting - ERP integration - Customer master data - Financial reporting - Automated commercial commitments


18. Glossary

Term Definition
INTUS The organization owning this project
GIGA CRM Multi-phase CRM roadmap name
AI Opportunity Platform Phase 1 product name
Opportunity Potential construction project business opportunity
Researcher User responsible for opportunity background research
Scheduler User responsible for meeting coordination, reminders, follow-up, and scheduling workflows
Enrichment Process of supplementing and completing opportunity information
Qualification Process of evaluating opportunity qualification context
Product Fit Evaluation of opportunity-product match
Confidence Indicator AI output confidence level indicator
Recommended Next Actions AI-suggested next actions
Handoff Package Structured information package from Researcher to Scheduler/Sales
Human Review Gate Required review step before customer-facing communication
Agent-assisted AI assists mode, AI operates within boundaries, provides suggestions, human confirms
Agent-operated AI autonomous execution mode, AI completes and notifies human
Human-gated Human gating mode, AI prepares, human approval required
Agent Governance Agent lifecycle, behavior, and safety management
System of Record Authoritative data source
Dodge Construction project information data service provider
ConstructConnect (CC) Construction project information data service provider
RBAC Role-Based Access Control
SOX Sarbanes-Oxley Act
RTO Recovery Time Objective
RPO Recovery Point Objective
UAT User Acceptance Testing
WBS Work Breakdown Structure

19. Appendix

Appendix A: Charter Summary

Appendix B: Open Items

ID Open Item Owner Due Date Status
O-01 Dodge API endpoint and authentication method β€” β€” TBD
O-02 ConstructConnect API endpoint and authentication method β€” β€” TBD
O-03 SSO system type and protocol β€” β€” TBD
O-04 AI model selection (cloud vs. on-premise) β€” β€” TBD
O-05 Deployment environment (cloud / on-premise) β€” β€” TBD
O-06 Pilot user list and scale β€” β€” TBD
O-07 Frontend tech stack selection β€” β€” TBD
O-08 Charter Rev.1 shortened timeline specifics β€” β€” TBD
O-09 Customer-facing communication policy approval (Constraint C-04) β€” β€” TBD
O-10 Relationship with existing CRM (push/standalone/bi-directional sync) β€” β€” TBD

Appendix C: Requirements Traceability Matrix Template

Req ID Description Source (Charter Clause) Priority Status Design Doc Dev Task Test Case
F-01-01 Platform workspace initialization Charter-Deliverable D-01 P0 Draft TBD TBD TBD
F-02-01 Dodge data import Charter-Deliverable D-02 P0 Draft TBD TBD TBD
... ... ... ... ... ... ... ...

End of Document

This PRD is Draft v0.2, revised based on verbatim reading of the charter images. It requires the following process before finalization: 1. Business review β€” Confirm scope and priorities 2. Architect review β€” Confirm technical feasibility 3. Development team review β€” Confirm implementation approach 4. Project manager approval β€” Confirm schedule and resources 5. Final release β€” v1.0 Baseline


Generated from GIGA_CRM_PRD_Framework_v0.1_EN.md — Deployed on Cloudflare Pages