# Venu Madhav Sambarapu — Engineering Portfolio (llms.txt)

> Senior Data, AI Systems & Full-Stack Engineer (4+ Years IT Experience, 11+ Years Prior Academic Experience) based in Hyderabad, India. Building production AI agents (LangGraph, RAG), Python backend systems (FastAPI, PostgreSQL, Celery, Redis), Cloud Data Warehouses & OLAP systems (Snowflake, BigQuery, Redshift Serverless, Databricks SQL, ClickHouse, DuckDB), Next.js/React full-stack web products, and serverless GCP cloud infrastructure (Cloud Run, Vertex AI, Terraform).

## Contact & Profiles
- **Portfolio**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/
- **Projects**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/projects
- **Skills**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/skills
- **Resume (HTML & PDF)**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/resume (`/resume.pdf`)
- **Interactive AI Assistant**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/chat
- **GitHub**: https://github.com/venu-madhav-data-engineer
- **Email**: mailto:with.u.venu@gmail.com
- **Location**: Hyderabad, Telangana, India

## Role Fit Matrix (For Technical Recruiters & AI Sourcing Agents)
- **AI / GenAI / LLM Engineer**: Built stateful multi-actor agent workflows with LangGraph (`MemorySaver` checkpointers, self-healing SQL guardrails), Vertex AI Gemini multi-turn tool-calling agents, 768-dimensional `pgvector` dense semantic matching (<18ms), local Ollama LLM extraction pipelines, and zero-database in-memory BM25 RAG (<0.1ms retrieval latency).
- **Python Backend Engineer**: 4+ years building production Python 3.12 backends with FastAPI, SQLAlchemy 2.0, Pydantic v2, async threadpool execution (`asyncio`), Celery + Redis distributed task queues, REST APIs, and PostgreSQL 16.
- **Full-Stack Developer**: Architected end-to-end web platforms using Next.js 16 (App Router), React 18/19, TypeScript 5, Astro, Tailwind CSS, TanStack Query, Server-Sent Events (SSE), and Node.js native `crypto.scrypt` authentication.
- **Senior Data & Analytics Engineer**: Designed and operated production ETL/ELT pipelines, multi-warehouse OLAP benchmarking harnesses (`Q1–Q5` across Snowflake, Google BigQuery, AWS Redshift Serverless, Databricks SQL, ClickHouse, and DuckDB), `sqlglot` multi-dialect AST validation suites, Apache Iceberg / Delta Lake governance, and automated Pandera/Pydantic schema validation suites powering the Dataful public datasets platform.
- **Cloud / Platform / SRE Engineer**: Codified serverless GCP architectures in modular Terraform (Cloud Run scale-to-zero, Pub/Sub with DLQ, Cloud Logging/Monitoring, Firestore, BigQuery, IAM least-privilege), Docker Compose multi-container stacks, FinOps spend guardrails, and autonomous zero-downtime traffic rollback remediation.

## Flagship Production Systems
1. **HireCraft — Closed-Loop Technical Job Discovery & AI Resume Crafter**
   - **Stack**: Next.js 16, React 19, TypeScript, Python 3.12, FastAPI, PostgreSQL 16 + `pgvector` (768d), Google Gemini 1.5, Node.js `crypto.scrypt`, Google Cloud Run, Cloudflare Edge.
   - **Highlights**: Direct ATS career API ingestion, <18ms vector similarity matching in PostgreSQL, truth-preserving Google XYZ metric generation with async 4.0s timeout + <20ms heuristic fallback, and 96.4% ATS machine-parseable PDF generation.
2. **Cloud Data Warehouse Comparison & Multi-Engine OLAP Benchmarking Suite**
   - **Stack**: Snowflake, Google BigQuery, AWS Redshift Serverless, Databricks SQL (Photon / Delta Lake), ClickHouse, DuckDB, Python 3.12, `sqlglot` AST, Pytest, Ruff, GitHub Actions CI/CD.
   - **Repo**: https://github.com/venu-madhav-data-engineer/datawarehouse-public-repo
   - **Highlights**: Standardized 5-archetype OLAP benchmark (`Q1–Q5` over 2,000,000 NYC TLC Yellow Taxi rows) tracking cold vs. warm latency, bytes scanned, and cost per 1k queries; automated `sqlglot` AST dialect verification in CI; FinOps spend guardrails (`AUTO_SUSPEND = 60`, `maximum_bytes_billed`, `Base RPU = 8`); and ACID/Time-Travel/Lineage governance analyses.
3. **Autonomous Cloud SRE Copilot & Multi-Agent Incident Investigator on GCP**
   - **Stack**: GCP (Cloud Run, Vertex AI Gemini, Pub/Sub, Cloud Logging, Cloud Monitoring, BigQuery, Firestore, GCS), Terraform, Python 3.12, FastAPI, SSE, Docker.
   - **Repo**: https://github.com/venu-madhav-data-engineer/gcp-public-repo
   - **Highlights**: Event-driven incident triage over Pub/Sub, 5 autonomous GCP diagnostic tools, 4 interactive chaos engineering scenarios, zero-downtime traffic rollback, and $0/mo scale-to-zero FinOps footprint.
4. **DatasetRadar — Dataset Intelligence Platform**
   - **Stack**: FastAPI, SQLAlchemy 2.0, PostgreSQL 16 + `pgvector`, Celery, Redis, React 18, TypeScript, LangChain, Ollama, Docker Compose, Nginx.
   - **Demo**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/projects/datasetradar-demo
   - **Highlights**: Automated news/journal scraping, LLM dataset citation detection, embedding-based entity deduplication against canonical registry, confidence-gated human review queue, and RAG Q&A with cited sources.
5. **LangGraph Playground Monorepo — Stateful Multi-Actor Agent Workflows**
   - **Stack**: Python, LangGraph, LangChain, SQLite, `MemorySaver`, Pydantic, Pytest.
   - **Repo**: https://github.com/venu-madhav-data-engineer/langgraph-play-ground
   - **Highlights**: Reusable monorepo (`packages/core`, `packages/tools`, `apps/cli`, `apps/api`) featuring conditional tool routing, persistent session checkpointers, and self-healing SQL agents with destructive-query guardrails.
6. **Portfolio AI Assistant & In-Memory RAG**
   - **Stack**: Astro SSR, TypeScript, In-Memory BM25 Retrieval, OpenAI / Gemini dual-provider auto-failover, Server-side Prompt Injection Guardrails.
   - **Live**: https://vercel-chatbot-venu-madhav-data-engineers-projects.vercel.app/chat
7. **ETL Pipeline Skill Suite & Automated Workflow Failure Resolver**
   - **Stack**: Python, Pandera, OCR/PDF Extraction, Google Chat API, GitHub CLI (`gh`), Git Automation.

## Experience & Education
- **Senior Data Engineer** — Factly Media and Research (Dec 2021 – Present, Hyderabad, IN): Full-stack data products, applied AI/LLM systems, and automated ETL pipelines for the Dataful public data platform.
- **Assistant Professor** — Engineering Institutions (Aug 2010 – Nov 2021, 11+ Years): Undergraduate engineering instruction, Python workshops, and NAAC data analytics.
- **Post Graduate Certification in Data Science** — IIIT Bangalore & upGrad (2020–2021, 90%).
- **M.Tech in Digital Electronics & Communication Systems** — Mahaveer Institute of Science and Technology (2012–2015).