Language: Go Source: https://arxiv.org/abs/2603.20534 Estimated lines: 6500
Manufacturing and regulated-industry engineering teams waste weeks manually extracting, cross-referencing, and auditing requirements across heterogeneous specification documents, supplier qualifications, and compliance standards — with no traceability and high error rates.
ReqTrace is a single Go binary that ingests industrial documents (PDFs, DOCX, spreadsheets), builds a hybrid semantic-lexical index, and exposes a local API and CLI for requirement extraction, cross-document traceability linking, compliance gap detection, and supplier risk scoring. Its key innovation is combining multi-provider LLM orchestration (routing queries to the cheapest adequate model) with a persistent traceability graph that tracks requirement evolution over time and flags drift, deletions, and emerging focus areas like IT security — turning the longitudinal analysis from the primary paper into a continuous monitoring capability rather than a one-off study.
["main.go","cmd/root.go","cmd/ingest.go","cmd/query.go","cmd/audit.go","cmd/serve.go","internal/ingest/parser.go","internal/ingest/pdf.go","internal/ingest/docx.go","internal/ingest/spreadsheet.go","internal/ingest/chunker.go","internal/index/hybrid.go","internal/index/semantic.go","internal/index/lexical.go","internal/index/store.go","internal/extract/requirements.go","internal/extract/entities.go","internal/extract/classifier.go","internal/trace/graph.go","internal/trace/linker.go","internal/trace/drift.go","internal/trace/timeline.go","internal/compliance/gap.go","internal/compliance/standards.go","internal/compliance/report.go","internal/supplier/risk.go","internal/supplier/scoring.go","internal/llm/router.go","internal/llm/provider.go","internal/llm/cost.go","internal/api/server.go","internal/api/handlers.go","internal/api/middleware.go","internal/config/config.go","internal/db/sqlite.go","internal/db/migrations.go","go.mod","go.sum","Dockerfile","reqtrace.example.yaml"]
This product is based on the following research papers. READ these to understand the technique you are implementing:
- 2603.20534
Existing repos that informed this design. STUDY these for prior art and patterns to improve on:
- infiniflow/ragflow
- langchain-ai/langchain
- langgenius/dify
- PaddlePaddle/PaddleOCR
- https://github.com/langgenius/dify
- https://github.com/langchain-ai/langchain
- https://github.com/open-webui/open-webui
- https://github.com/Shubhamsaboo/awesome-llm-apps
- https://github.com/infiniflow/ragflow
- https://github.com/PaddlePaddle/PaddleOCR
- https://github.com/dair-ai/Prompt-Engineering-Guide
The buyer is the requirements engineering or quality/compliance team lead at automotive OEMs, aerospace firms, and Tier-1 suppliers — companies already spending $50-200K/year on manual requirement management tools like IBM DOORS or Jama Connect. ReqTrace sells as an on-prem enterprise license ($2K-8K/seat/year) with a free single-user CLI tier for adoption, targeting the 83% time reduction and contract-penalty avoidance ($2.3M in the paper's case) as ROI justification. The moat is the traceability graph and longitudinal drift detection — general-purpose RAG platforms like Dify and RAGFlow lack domain-specific requirement linking, compliance gap scoring, and supplier risk models, while incumbent RE tools lack LLM-powered extraction entirely.