Skip to content
Farid Gurbanov

Portfolio

Working code behind the positioning, across three areas: AI & automation; integration, streaming & data; and applied ML & data science. Every card carries an honest maturity label — nothing is dressed up as more than it is.

Build proof, honestly labelled

AI & automation

  1. 1Govern & route the models

    sovereign-llm-gateway

    Working — runs end-to-end

    What it proves

    A working LLM gateway: per-agent cost and budget enforcement, vendor abstraction (LiteLLM), a local-model fallback (Ollama) for sovereignty, and Prometheus observability. Runs end-to-end with `docker compose up`.

  2. 2Build a trustworthy agentic product

    sovereign-copilot

    Reference architecture

    What it proves

    A reference architecture for a trustworthy copilot: deterministic tool contracts (MCP), retrieval grounded in your data (BGE-M3 + reranker), L1–L4 evaluation gates with goldens, and answers that trace to a recorded call chain.

  3. 3Deliver software with agents

    maestro

    Reference architectureMIT

    What it proves

    A reference architecture for spec-driven delivery: agents propose, humans dispose. Functional and technical gates enforced through GitHub branch protection across a multi-repo, multi-participant product.

Integration, streaming & data

  1. 1Stream & integrate events at platform scale

    event-integration-platform

    Working — runs end-to-endMIT

    What it proves

    A Kafka-native, multi-tenant event-streaming & integration platform: REST→Kafka ingest, managed JSONata transforms with DLQ + replay, Kafka Connect HTTP/S3 sinks, a control-plane API and a drag-and-drop pipeline UI, all under workspace-scoped observability. Runs locally with `docker compose up`.

    View repository

Applied ML & data science

  1. 1Detect contrails in sky-camera images

    contrail-segmentation-demo

    Working — runs end-to-end

    What it proves

    A neural-network image-segmentation app — React + TypeScript front end → Node.js (Express) BFF → Python FastAPI service → a hand-written PyTorch U-Net — that detects contrails in sky-camera images and reports coverage and contrail count. Three services that run end-to-end with `docker compose up`, with CI on GitHub Actions. Trained on a synthetic sky generator so it runs on a laptop in minutes; the README writes down the path to real GVCCS / Sky-Cam imagery.

    View repository
  2. 2Decide which flights to reroute — and at what cost

    contrail-avoidance-pipeline

    Working — runs end-to-end

    What it proves

    A Polars/Pandas pipeline plus a Databricks-style notebook that flag which flights form persistent, climate-warming contrails — using the Schmidt–Appleman Criterion and ice-supersaturated regions — and propose altitude changes, weighing avoided climate forcing (CO₂e) against extra fuel burn. Lakehouse-shaped (Parquet, Delta / Unity-Catalog framing) and runs end-to-end on a laptop, with a documented path to ERA5 reanalysis + OpenSky real flight data.

    View repository
  3. 3Price retail at scale — elasticity to optimization

    retail-dynamic-pricing

    Working — runs end-to-endMIT

    What it proves

    A retail dynamic-pricing platform on a Databricks lakehouse: one elasticity-to-optimization engine serving two verticals — grocery (elasticity & markdown) and consumer electronics (competitive & lifecycle, MAP-compliant). Log-log demand estimation checked against a known ground truth, a solver-agnostic revenue optimizer (SciPy by default, Gurobi MIQP backend), and Delta Live Tables / Workflow pipelines. Two notebooks run end-to-end on a laptop on synthetic data — +6.4% revenue at flat margin in the grocery run — with the path to production written down.

    View repository

The applied demos and the event-integration platform are public and link out. The remaining repos — the AI & automation trio — are being aligned under one name, neutralized and licensed before they carry my name on a public surface; their descriptions are honest now, links follow.

Five of these run end-to-end today — the LLM gateway, the event-integration platform, the retail dynamic-pricing demo and the two contrail demos. The other two are reference architectures, and each card says which it is. No production-system claim I can’t back.