Builder with 9+ years in Financial Services, specializing in Applied Agentic AI. My journey started as a software developer in 2003, moved through integration, application, and enterprise architecture — then GenAI hit in 2022 and I took a complete U-turn back to builder + product owner. I have never looked back.
- Architect — Applied AI & Agentic Systems, Freedom Mortgage / Xpanse (2022–present)
- Architect — Digital Platform (0-to-1), Freedom Mortgage (2019–2022)
- Solutions Architect — Customer Enablement, Salesforce (2018–2019)
- Integration Architect, Accenture — deployed at QBE North America (2016–2018)
- Integration Architect — Government Initiatives, MD Health Exchange / NJ Courts (2013–2016)
- Integration Architect, IBM — deployed at American Airlines & IBM Software Group (2004–2013)
- Product & Strategy: 0-to-1 Product Development, Hands-On Prototyping, Rapid POC Development, Roadmap Ownership, Vendor Evaluation, Stakeholder Alignment
- AI & Architecture: Agentic AI Solutions, LLM Deployment, MCP Servers & Tools, Neuro-Symbolic Architecture, RAG Pipelines, Evaluation Frameworks, Context Engineering
- Cloud & Enterprise: AWS (Lambda, API Gateway, Serverless), Snowflake, Salesforce, MuleSoft, Palantir Foundry/AIP, Okta/Azure AD
- Regulated Industries: Mortgage Origination & Servicing, Insurance Operations, Healthcare Exchanges, Compliance & Audit
- Master of Computer Applications (MCA) — West Bengal University of Technology, 2004
- B.Sc. Mathematics — Calcutta University, 2001
- Champion of Innovation (2024, 2022)
- Highest Flyer (2023)
A neuro-symbolic agentic harness for mortgage workflows — deterministic rules engines + LLM reasoning to automate loan origination decisions with full compliance auditability. Every decision recorded with lineage, building institutional knowledge like case law for underwriting.
AI-powered title commitment review for mortgage origination.
CLI tooling for managing rules in Palantir Foundry.
Model Context Protocol servers for AI agent integrations.
AI/ML learning community and podcast ("Demystifying Hype to Reality"), hosting 30+ industry sessions on applied AI, agent architectures, and enterprise deployment patterns.
AI will write all software. The most important thing we as humans will do is ideation, intent classification, and outcome measurement. The outcome is solving business problems with either profit increase or cost reduction.