Wiki source code of POC (Proof of Concept)
Last modified by Robert Schaub on 2026/02/08 08:26
Show last authors
| author | version | line-number | content |
|---|---|---|---|
| 1 | = POC (Proof of Concept) = | ||
| 2 | |||
| 3 | **Status:** ✅ Planning Complete - Ready for Development | ||
| 4 | **Version:** 1.0 | ||
| 5 | **Goal:** Prove that AI can extract claims and determine verdicts automatically without human intervention | ||
| 6 | |||
| 7 | ---- | ||
| 8 | |||
| 9 | == Overview == | ||
| 10 | |||
| 11 | POC1 tests the core FactHarbor capability: Can AI automatically extract factual claims from articles and evaluate them with reasonable verdicts? | ||
| 12 | |||
| 13 | **What POC Tests:** | ||
| 14 | |||
| 15 | * AI claim extraction from article text | ||
| 16 | * AI verdict generation with confidence scores | ||
| 17 | * Fully automated approach viability (no human editing) | ||
| 18 | * Output comprehensibility to users | ||
| 19 | |||
| 20 | **What POC Excludes (Intentional Simplifications):** | ||
| 21 | |||
| 22 | * Scenarios (deferred to POC2 - architectural questions remain) | ||
| 23 | * Evidence display (deferred) | ||
| 24 | * Production infrastructure (simplified single-call architecture) | ||
| 25 | * Full data model (stateless POC) | ||
| 26 | |||
| 27 | **Key Decision:** Test AI capability first, add architectural complexity in POC2 based on learnings. | ||
| 28 | |||
| 29 | ---- | ||
| 30 | |||
| 31 | == POC Documentation == | ||
| 32 | |||
| 33 | **POC-Specific Pages:** | ||
| 34 | |||
| 35 | * [[Requirements>>FactHarbor.Specification.POC.Requirements]] - Complete POC specification (17 sections, 1,197 lines) | ||
| 36 | * [[Summary>>FactHarbor.Specification.POC.Summary]] - Executive overview and consolidated analysis | ||
| 37 | |||
| 38 | ---- | ||
| 39 | |||
| 40 | == Related Specification Pages == | ||
| 41 | |||
| 42 | **Core Specification:** | ||
| 43 | |||
| 44 | * [[Requirements (Full System)>>Archive.FactHarbor 2026\.01\.20.Specification.Requirements.WebHome]] - FR1-FR13, NFR1-NFR5 | ||
| 45 | * [[User Needs>>Archive.FactHarbor 2026\.01\.20.Specification.Requirements.User Needs.WebHome]] - UN-1 through UN-17 | ||
| 46 | * [[Gap Analysis>>FactHarbor.Specification.Requirements.GapAnalysis]] - Feature gaps, proposes FR14-FR43 | ||
| 47 | |||
| 48 | **Architecture & Design:** | ||
| 49 | |||
| 50 | * [[Architecture>>Archive.FactHarbor 2026\.01\.20.Specification.Architecture.WebHome]] - System architecture and design | ||
| 51 | * [[AKEL>>Archive.FactHarbor 2026\.01\.20.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]] - AI subsystem | ||
| 52 | * [[Data Model>>Archive.FactHarbor 2026\.01\.20.Specification.Data Model.WebHome]] - Database schema | ||
| 53 | * [[Workflows>>Archive.FactHarbor 2026\.01\.20.Specification.Workflows.WebHome]] - Process workflows | ||
| 54 | |||
| 55 | ---- | ||
| 56 | |||
| 57 | == Key POC Specifications == | ||
| 58 | |||
| 59 | **Output Components (4):** | ||
| 60 | |||
| 61 | 1. Analysis Summary (3-5 sentences) | ||
| 62 | 2. Claims Identification (3-5 claims) | ||
| 63 | 3. Claims Verdicts (verdict + confidence + risk tier + reasoning per claim) | ||
| 64 | 4. Article Summary (optional, 3-5 sentences) | ||
| 65 | |||
| 66 | **Publication Mode:** Mode 2 (AI-Generated, No Prior Human Review) | ||
| 67 | |||
| 68 | **Quality Gates:** 4 simplified gates (Source Quality, Contradiction Search, Uncertainty, Structural Integrity) | ||
| 69 | |||
| 70 | **Success Criteria:** ≥70% AI quality without manual editing | ||
| 71 | |||
| 72 | ---- | ||
| 73 | |||
| 74 | == POC vs. Full System == | ||
| 75 | |||
| 76 | |=Aspect|=POC1|=Full System | ||
| 77 | |Scenarios|None (deferred to POC2)|Core component (Claims → Scenarios → Verdicts) | ||
| 78 | |Workflow|3 steps|6 phases with quality gates | ||
| 79 | |AKEL|Single API call|Multi-component orchestrated pipeline | ||
| 80 | |Data|Stateless (no DB)|PostgreSQL + Redis + S3 | ||
| 81 | |Architecture|Simplified proof-of-concept|Production architecture | ||
| 82 | |Requirements|FR-POC-1 to FR-POC-7|FR1-FR43 (full system) | ||
| 83 | |||
| 84 | **Gap Between POC and Beta:** Significant architectural expansion needed (scenarios, components, data model). | ||
| 85 | |||
| 86 | ---- | ||
| 87 | |||
| 88 | == Decision Gate == | ||
| 89 | |||
| 90 | After POC testing complete, team makes one of three decisions: | ||
| 91 | |||
| 92 | **GO (Proceed to POC2):** AI quality ≥70%, approach validated | ||
| 93 | **NO-GO (Pivot or Stop):** AI quality < 60%, fundamental flaws identified | ||
| 94 | **ITERATE (Improve POC):** AI quality 60-70%, specific improvements needed | ||
| 95 | |||
| 96 | ---- | ||
| 97 | |||
| 98 | **Next Steps:** | ||
| 99 | |||
| 100 | 1. Review [[Requirements>>FactHarbor.Specification.POC.Requirements]] for detailed specifications | ||
| 101 | 2. Review [[Gap Analysis>>FactHarbor.Specification.Requirements.GapAnalysis]] for feature priorities | ||
| 102 | 3. Begin POC development | ||
| 103 | 4. Test with diverse articles | ||
| 104 | 5. Make GO/NO-GO decision | ||
| 105 | |||
| 106 | **Document Version:** 2.0 (Updated after Specification Cross-Check) |