Finance and operations knowledge. Production AI agents. Real-time computer vision.
Bengaluru, India · Email me
I build AI systems for people who need dependable answers from business data, documents and the physical world.
My experience connects financial analytics, enterprise software and real-time machine learning. I work with business stakeholders to understand the problem, build the system and check that its outputs hold up in production.
| Combination | What that means in practice |
|---|---|
| Finance + AI agents | Turn questions about margins, costs, inventory and working capital into analytical tools and AI-assisted answers. |
| Enterprise data + answer validation | Work with SAP data, SQL and document retrieval; use source reconciliation, regression tests and tracing to check results. |
| Computer vision + production ML | Build retail recognition and industrial inspection systems, optimize inference and deploy model-serving APIs. |
| Customer delivery + cloud engineering | Translate business requirements into architecture, implementation and deployment, with expertise across AWS, Azure and GCP. |
- Enterprise financial analytics: Built LangGraph orchestration over SAP-backed analytical tools and traced financial figures from source calculations to user-facing answers.
- Process knowledge: Turned business-process recordings into reusable documentation; built asynchronous ingestion, hybrid vector + SQL retrieval and persistent conversation history.
- Real-time retail vision: Delivered multi-camera inference with sub-200 ms response times at 15–20 FPS, using YOLO, TensorRT and LightGlue.
- Operational intelligence: Built RFID model-serving APIs and LLM-assisted log investigation workflows supporting 1,000+ review cycles.
Most recent enterprise work lives in private repositories. The public repositories here mainly reflect earlier independent and academic work; my portfolio provides broader context on my experience.
| Area | Technologies |
|---|---|
| Agents, retrieval & evaluation | LangGraph, LangChain, RAG, pgvector, Langfuse |
| Software & data | Python, SQL, Go, C++, FastAPI, PostgreSQL, Redis, MongoDB |
| Cloud & deployment | AWS, Azure, GCP, Docker, Kubernetes, CI/CD |
| Vision & edge | TensorRT, ONNX, OpenCV, YOLO, LightGlue |
| Enterprise integration | SAP Datasphere, OData, analytical APIs, KPI reconciliation |
Currently Senior Integration Consultant — Analytics & Information Management at Cervello India, a Kearney company, focused on Applied AI / GenAI engineering. Previously built AI and ML systems at SOLUM and Ignitarium. I also mentor two engineers and conduct AI engineering interviews.
- M.Tech, Data Science · B.Tech, Mechanical Engineering — Amrita School of Engineering
- Google Cloud Certified — Professional Cloud Architect
- Anthropic Claude Certified Architect — Professional
Interested in Senior Applied AI and Forward Deployed Engineering roles that support remote work from India, particularly enterprise agents, finance and operations, and reliable AI platforms.
Portfolio & Finance × AI writing · LinkedIn · phvpavankumar@gmail.com