Skip to content

utkanbir/Semantic-Intelligence-Platform

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

372 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Semantic Intelligence Platform (SIP)

Turn enterprise data into governed, explainable semantic intelligence — from discovery to agents.

SIP is a modular, cloud-native platform that guides organizations through the full Semantic Intelligence lifecycle: structured discovery, blueprint-driven provisioning, curated data products, ontology-backed knowledge graphs, and policy-governed AI agents. Every significant action is traceable via Semantic Transactions and Trace Steps, so outcomes stay auditable and explainable by design.

The current develop branch includes a working Platform Console for Applications, Connectors, a unified Ontology Creation Wizard (Manual / Import / Generate), LLM advisory semantic review, and Semantic Transaction inspection on Kubernetes (console.sip.local, api.sip.local).


Why SIP?

Modern data platforms often ship disconnected pipelines and opaque AI wrappers. SIP takes a different path:

Principle What it means in practice
Blueprint before provisioning Nothing is provisioned blindly — a reviewed Blueprint defines what gets built
Products, not raw tables Agents and consumers work with Published Data Products, not ad-hoc SQL
Explainability by construction Discovery, provisioning, and agent runs emit semantic audit trails
Technology independence Business logic talks to ports (storage, KG, vectors, LLM) — not vendor SDKs
Governed delivery Architecture-first sprints, ADRs, and sprint-close quality gates

Platform lifecycle

flowchart LR
    A[Applications] --> B[Discovery]
    B --> C[Blueprints]
    C --> D[Assets & Workspace]
    D --> E[Data Products]
    E --> F[Ontology & KG]
    F --> G[Agents]
    G --> H[Agent Runtime]
    P[Governance] -.-> B & C & E & G
    T[Audit Trace] -.-> B & C & D & E & G & H
Loading

What's in the monorepo

Path Purpose
backend/ Python / FastAPI modular monolith — REST API at /api/v1
frontend/ React / TypeScript Platform Console — applications, connectors, ontology wizard, semantic transactions
infra/ Kubernetes + Kustomize — primary runtime (sip-dev namespace)
architecture/ Canonical architecture specifications
docs/ ADRs, governance, retros, and engineering playbooks
scripts/ Sprint verification, board sync, and approved automation

Backend modules (MVP)

applications · discovery · blueprints · assets · audit_trace · products · agents · ontology · knowledge_graph · adapters · agent_runtime · governance


Tech stack

  • API: FastAPI, Pydantic, Alembic (PostgreSQL)
  • Architecture: Modular monolith, Ports & Adapters (hexagonal)
  • Runtime: Kubernetes-first (ADR-001); local dev via sip-dev + port-forward
  • Integrations (via adapters): PostgreSQL, MinIO, Fuseki, Qdrant, OpenMetadata, OpenAI

Quick start

# Clone and enter the repo
git clone https://github.com/utkanbir/Semantic-Intelligence-Platform.git
cd Semantic-Intelligence-Platform

# Backend (see backend/README.md for SIP_* env vars)
cd backend
pip install -e ".[dev]"
alembic upgrade head
uvicorn app.main:app --reload

# Kubernetes dev cluster — see infra/README.md
kubectl apply -k infra/kubernetes/overlays/dev

Health check: GET /api/v1/health

Console ingress: http://console.sip.local
API ingress: http://api.sip.local


Documentation

Document Description
Project handoff Current state, live env, onboarding checklist
Software architecture guide Backend/frontend layout, modules, ports, layering
Architecture supplements index Per-domain contracts + staleness notes
Development Playbook Engineering process, quality gates, team conventions
GitHub Workflow Issues, sprints, milestones, delivery workflow
ADR-001: Cloud Native Deployment Kubernetes-first runtime strategy
Governance Decision authority, architecture gates, sprint retros
Backend README API modules, configuration, migrations
Infra README Cluster setup, deploy, port-forward

Project status

SIP MVP is under active sprint delivery on the develop integration branch.

Milestone Status
Foundation & Applications (Sprint 0–1) ✅ Closed
Discovery & Blueprint (Sprint 2–3) ✅ Closed
Assets & Audit Trace (Sprint 4) ✅ Closed
Products & Agents (Sprint 5–6) ✅ Closed
Ontology & Knowledge Graph (Sprint 7) ✅ Closed
Adapters & Agent Runtime (Sprint 8) ✅ Closed
Governance (Sprint 9) ✅ Closed
Assessment MVP E2E (Sprint 10) ✅ Closed
Platform Console (Sprint 11) ✅ Closed
Console Deploy & Release (Sprint 12) ✅ Closed
Console Connectors UX v2 (Sprint 29) ✅ Closed
Fuseki Persistence & Semantic Transactions (Sprint 30) ✅ Closed
Ontology Wizard & Semantic Transactions v2 (Sprint 31) ✅ Closed
Semantic Transaction Realignment (Sprint 32) ✅ Closed
Ontology Creation Flow v2 (Sprint 34) ✅ Closed
Ontology Flow Polish (Sprint 35) ✅ Closed

Current: Default branch develop · Console on K8s (console.sip.local) · Unified ontology wizard (3 modes, draft-first) · LLM advisory review UI · Alembic 20260706_0019 · Images sip-backend:s53 / sip-console:s54

Recent delivery highlights

  • Ontology Creation Wizard (Sprint 34): Manual forms, OWL/RDF import (file + paste), Generate from Sources; shared Validate → Connector → Review → Approve → Materialize lifecycle; graph write only after explicit approval.
  • Advisory semantic review (Sprint 34–35): Structured advisory findings (suggestions, warnings, improvements) after deterministic validation; Accept/Ignore per suggestion; never auto-mutates the draft. Uses stub LLM port until a live provider is configured.
  • Generate extensions (Sprint 35): Web URL fetch, CSV/Excel client parse, relaxed connector gate for generate-only draft phase.
  • Unified Connectors UX: Platform hub, list-first connector registry, Fuseki persistence on materialize.
  • Semantic Transactions UI: Dedicated semantic lineage surface separate from operational audit trace.

Contributing

Work is tracked in GitHub Issues and the SIP MVP Delivery project board. See SIP GitHub Workflow for issue templates, sprint milestones, and PR standards. Architecture changes require ADR review — implementation follows frozen MVP specs in architecture/.


License

See repository license file for terms.

About

Governed semantic intelligence platform — discovery, blueprints, data products, knowledge graphs, and explainable AI agents. FastAPI modular monolith on Kubernetes.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Packages

 
 
 

Contributors