1. Roadmap Principles
- Benefit-first ordering: highest commercial value delivered earliest
- Dependency-aware sequencing: no work package starts before its prerequisites complete
- Farm-out ready: each package has clear scope, inputs, outputs, and acceptance criteria
- Incremental delivery: each phase produces deployable, demonstrable capability
2. Dependency Graph
Formula Parser (DONE)
|
+-> Rule Engine (DONE)
| +-> EDR Pipeline (DONE) -----> EDR Use Case (Phase 2)
| +-> PFP Pipeline (DONE) -----> Nesto Use Case (Phase 1)
| +-> Freelance Pipeline (DONE)
| +-> TCO Pipeline (DONE)
|
+-> BDT Validator (DONE)
| +-> ValidateNode (DONE) -----> Form+Validate Pattern (Phase 1)
|
+-> Orchestration Engine (DONE)
|
+-> Node Executors (18/29 DONE)
| +-> Scaffold Nodes (10 remaining) -----> Phase 2-3
|
+-> deltaPrism GPU Runtime (DONE)
| +-> GPU Batch Processing -----> Phase 3 (AES)
|
+-> Runtimes
|
+-> Local (DONE)
+-> Databricks (SCAFFOLD) -----> Phase 3
+-> Snowflake (SCAFFOLD) -----> Phase 3
+-> PowerFx Transpiler (DECLARED) -----> Phase 4
3. Phase 1: Nesto MVP (Highest Benefit)
Timeline: 4-6 weeks per cooperator
Commercial value: First paying use case, UAE compliance market
Dependencies: All satisfied (Rule Engine, BDT, ValidateNode, FormNode, Orchestration Engine)
| Work Package | Assignable | Inputs | Outputs | Acceptance |
|---|---|---|---|---|
| WP-1.1 Nesto Flow Assembly | Yes | 17 compliance rules, FormNode, ValidateNode, HumanGate | Complete Nesto pipeline JSON | All 17 rules evaluate correctly in flow |
| WP-1.2 HumanGate Webhook DONE | Yes | HumanGateNode scaffold (90 lines) | Production approval callback | Approval pauses flow, resumes on webhook |
| WP-1.3 Nesto ChameleonV2 View | Yes | Entity schema, V-1.0 contract | React forms + dashboard | All 17 fields render with BDT validation |
| WP-1.4 Email Notification DONE | Yes | EmailForward scaffold (108 lines) | SMTP integration | Approval/rejection emails send correctly |
| WP-1.5 Nesto NSD Complete DONE | Yes | All above outputs | Versioned NSD JSON | Passes conformance CT-RULE, CT-BDT, CT-FLOW |
Benefit: Deployable compliance tool for UAE worker management. Revenue-generating.
4. Phase 2: EDR Production + ARC Lead Pipeline
Timeline: 6-8 weeks per cooperator
Commercial value: Pyrometrix/oil-gas engineering data validation; ARC CRM automation
Dependencies: Phase 1 HumanGate webhook (shared component)
| Work Package | Assignable | Inputs | Outputs | Acceptance |
|---|---|---|---|---|
| WP-2.1 EDR Entity Classes DONE | Yes | 6 entity classes in arc/entity_classes.py | Complete BDT-typed entity schemas | All fields mapped to BDT types |
| WP-2.2 EDR Validation Rules DONE | Yes | 19 edr_validation rules | Extended rule set with edge cases | 95%+ coverage on industrial entity types |
| WP-2.3 ARC LeadScore Tuning DONE | Yes | LeadScoreNode (159 lines) | Calibrated scoring model | Score correlates with conversion rate |
| WP-2.4 ARC MatchNode Production DONE | Yes | MatchNode (179 lines) | Fuzzy matching with threshold config | De-duplication rate >90% on test data |
| WP-2.5 AtlasOCR Integration DONE | Yes | AtlasOcr scaffold (112 lines) | Connected OCR API | Field extraction from P&ID drawings |
| WP-2.6 PDF Report Generator DONE | Yes | PdfGenerator scaffold (93 lines) | Template-driven PDF output | EDR compliance reports generate correctly |
Benefit: Engineering data quality tool (Pyrometrix clients), CRM automation (ARC clients).
5. Phase 3: AES + GPU Batch + Advanced Runtimes
Timeline: 8-12 weeks
Commercial value: Assessment engine for education/certification; enterprise batch processing
Dependencies: Phase 2 EDR entity patterns, deltaPrism GPU runtime
| Work Package | Assignable | Inputs | Outputs | Acceptance |
|---|---|---|---|---|
| WP-3.1 AES Assessment Engine DONE | Yes | FormNode, AiTransform, ConditionBranch, DashboardNode | Assessment flow with AI grading | Multi-step assessment with branching |
| WP-3.2 GPU Batch Scaling DONE | Yes | deltaPrism GPU runtime (963 lines) | 100K+ records/minute throughput | Benchmark passes on test dataset |
| WP-3.3 Databricks Runtime DONE | Yes | Runtime scaffold (418 lines) | PySpark rule evaluation | CT-RULE, CT-FORM pass on Databricks |
| WP-3.4 Snowflake Runtime DONE | Yes | Runtime scaffold (231 lines) | Snowpark UDF evaluation | CT-RULE, CT-FORM pass on Snowflake |
| WP-3.5 ChameleonBI (Osiris) DONE | Yes | Osiris prototype research | BI dashboard component for Chameleon | Real-time data visualisation, drill-down |
| WP-3.6 SMS Gateway Integration DONE | Yes | SmsIngest + SmsReply scaffolds | Twilio/MessageBird integration | Send/receive SMS in flows |
Benefit: Assessment SaaS vertical, enterprise-scale processing, cloud-native deployments.
6. Phase 4: Platform Expansion
Timeline: 12-16 weeks
Commercial value: Microsoft enterprise market, legacy system integration
Dependencies: Phase 1-3 core patterns proven
| Work Package | Assignable | Inputs | Outputs | Acceptance |
|---|---|---|---|---|
| WP-4.1 PowerFx Transpiler DONE | Yes | Formula Parser (600 lines) | PowerFx formula output | All 11 functions translate correctly |
| WP-4.2 Power Automate Connector DONE | Yes | Orchestration Engine patterns | Power Automate flow export | Nesto flow runs in Power Automate |
| WP-4.3 SharePoint Model DONE | Yes | Domain model, M-1.0 contract | SharePoint list schemas | CRUD operations pass CT-LIFE |
| WP-4.4 VBA Macro Library DONE | Yes | Formula Parser | VBA evaluation functions | Excel prototype deployment works |
| WP-4.5 COBOL Copybook Generator DONE | Yes | Formula Parser, BDT | COBOL evaluation logic | Mainframe deployment feasible |
Benefit: Microsoft 365 market access, legacy banking integration.
7. Critical Path Analysis
The critical path (longest dependency chain to full platform):
Formula Parser -> Rule Engine -> Orchestration Engine -> HumanGate Webhook
-> Nesto MVP -> EDR Production -> AES Engine -> Databricks Runtime
-> PowerFx Transpiler
Path of most benefit (highest ROI per unit effort):
- WP-1.2 HumanGate Webhook -- unblocks ALL approval flows across ALL use cases
- WP-1.5 Nesto NSD -- first revenue-generating deployment
- WP-2.5 AtlasOCR -- unblocks document processing for EDR + ENDO
- WP-3.1 AES Assessment -- new vertical market
- WP-4.1 PowerFx Transpiler -- unlocks Microsoft enterprise market
8. Cooperator Assignment Matrix
| Cooperator Skill | Suitable Work Packages |
|---|---|
| Python backend + async | WP-1.2, WP-1.4, WP-2.4, WP-3.2, WP-3.6 |
| React + TypeScript | WP-1.3, WP-3.5 |
| Data engineering (Spark/Snowflake) | WP-3.3, WP-3.4 |
| AI/ML integration | WP-2.3, WP-2.5, WP-3.1 |
| Microsoft 365 / Power Platform | WP-4.1, WP-4.2, WP-4.3 |
| Document/PDF processing | WP-2.6, WP-3.1 |
| Legacy systems (VBA/COBOL) | WP-4.4, WP-4.5 |
| Domain expert (compliance/UAE) | WP-1.1, WP-1.5, WP-2.1, WP-2.2 |
9. Version History
| Version | Date | Changes |
|---|---|---|
| RM-1.0 | 2026-05-02 | Initial roadmap: 4 phases, 22 work packages, dependency graph, benefit ordering. |