Headline

Founder · Safe AI systems for regulated industries · Data Engineer · Ex-Siemens

About

I build systems that take the grunt work off people's plates, so they can spend time on the things that actually require them. Most knowledge work has two layers. There's the layer that requires genuine expertise: judgement, creativity, client relationships, knowing which questions to ask. And there's the layer underneath, reading, extracting, cross-referencing, formatting, that consumes most of the time but almost none of the thinking. That gap is what I'm building toward closing. At Craton Labs, that means document intelligence and workflow tooling for consulting firms, where the regulatory document volume is high, the grunt work is well-defined, and the people doing it are too good to be doing it. On the engineering side: RAG pipelines, LLM orchestration, document intelligence, citation verification, multi-tenant architecture on AWS. Previously: rail analytics at Siemens, edge AI at Habitat, data engineering at an insurtech. Open to AI engineering consulting alongside Craton Labs. Based in Bangalore; available Delhi, Mumbai.

Location

India

AI stack

antigravity

claudecode

Work experience

Founder at Stealth Startup, Oct 2025 - Present · 11 mos, India, Building Atlas: a workspace for BFSI compliance consultants. The problem: a boutique compliance firm spends the majority of its time reading regulatory documents, extracting obligations, and building matrices manually. It's high-volume, low-creativity work that consumes the time of people whose real value is judgement and client relationships. Atlas automates that layer. A consultant uploads a regulatory circular and a client policy document. The system parses, classifies, extracts obligations, verifies citations against source text, identifies gaps, and generates a deliverable-ready compliance matrix, with a full audit trail of what it found and why.

Data Engineer at RiskPe Consultants, Jan 2025 - Sep 2025 · 9 mos, Jaipur, Built the data infrastructure that ran daily operations at an early-stage insurtech. PySpark ETL pipelines on AWS EC2 processing 10K+ daily policy and lead records; reduced reporting latency by 40% Power BI dashboards for underwriting, fraud detection, and conversion tracking across five business units Prophet-based forecasting models improving sales planning accuracy by 12–15% Fraud signal detection using behavioral and transactional patterns, dataset and feature logic later fed into an ML-ready pipeline for the underwriting team Stack: AWS EC2 · PySpark · Power BI · Prophet · SQL

Edge AI Engineer at Habitat (Sheshuki Automation Pvt. Ltd.), Jul 2024 - Dec 2024 · 6 mos, Bengaluru, Deployed a YOLOv7-based ANPR system on BeagleY-AI edge hardware , replacing a proprietary system at roughly half the cost. The constraint: low-power ARM, no cloud dependency, real-time processing required Achieved 15+ FPS through OpenCV pipeline optimization and model compression Built edge-to-cloud telemetry layer for centralized fleet monitoring Stack: YOLOv7 · OpenCV · BeagleY-AI · ARM · Edge-to-cloud architecture

Personal goal pursuit at Career Break, Jul 2023 - Jun 2024 · 1 yr, Jaipur, Rajasthan, Took time off to up skill myself, solidify my goals and equip myself to better pursue them.

System Engineer at Siemens, Jul 2022 - Jun 2023 · 1 yr, Bengaluru, Rail analytics across Vienna Metro, Sydney Airport Line, and Indian Railways projects. - Python and Power BI pipelines for condition monitoring of rolling stock sensor data - Automated manual processing workflows, cutting effort by ~30% - Failure prediction models for proactive maintenance scheduling - Telemetry PoC for low-connectivity rail corridors, early exposure to making data reliable in environments where infrastructure can't be assumed Stack: Python · Power BI · SQL · ETL · IoT (ESP32)

Bachelor Student - Technical at Siemens, Jul 2021 - Jun 2022 · 1 yr, India

Education

Master's Degree, Data Science and AI, Birla Institute of Technology and Science, Pilani

Diploma (Online), Data Science and Programming, Indian Institute of Technology, Madras