SHIPPING PRODUCT SYSTEMS

Product leader who ships. Grounded in actual testing.

18+ years in product across US Retirement, Insurtech, and AI/ML, turning enterprise platforms into shipped, working systems. I don't just spec AI features — I build working versions first, ensuring every product call I make is validated firsthand.

v1.2.0

FitCheck

Live

An AI-powered job-fit analyzer for senior professionals, designed, built, and deployed entirely as a solo project. Scores profile alignment, flags skill gaps, identifies system red-flags, and suggests contextual improvements.

Privacy Standards

Built natively to DPDP 2023 design patterns, featuring a strict user-consent interface workflow and an entirely stateless, zero-data-retention server architecture.

View live application console →
Tech Stack
PythonFlaskGroq APILlama 3.3 LayerRender
v1.0.0

Multi-LLM Model Setup

Live

A self-directed exploration of model orchestration boundaries. Seamlessly links disparate local configurations into a centralized runtime desktop client as context-aware, callable tools.

Operational Logic

Utilizes custom Model Context Protocol scripting to bridge specialized engines. Offloads analytical processing steps dynamically based on workload complexity at zero overhead.

Tech Stack
Ollama BaseQwen 2.5 (7B)Llama 3.2 (3B)FastMCPPython Core
v0.4.0

DharaBot

In Development

Bypasses typical crop disease diagnostic delays that cost structural yield, providing remote workforces sub-second agricultural clarity through basic edge photography inputs.

Core Architecture

Bilingual processing layer parsing lightweight payloads. Leverages localized edge worker nodes to avoid enterprise single points of failure, optimized natively for mobile caching layers.

Tech Stack
Cloudflare Workers AIRedis Edge CacheRender ProcessingStateless Architecture

Hands-on AI Prototyping

Bypassing superficial feature specs by building actual technical models first. Deeply understanding architectural tradeoffs, local compute configurations versus cloud APIs, and scaling realities before handing work over to engineering teams.

Regulated Domain Scale

Navigating structural transformation inside highly constrained environments. 18+ years turning complex processes across US Retirement, Insurtech, and Financial platforms into stable, resilient software ecosystems.

Team Building & Discovery

Assembling core groups from scratch and guiding cross-functional teams through ambiguous engineering sprints. Translating emerging research (like biometrics and affective computing MVPs) into real-world software products.

Rigor & Backlog Ownership

Bridging stakeholder demands with clear, structured execution. Running process workshops, managing agile delivery tracks, and proactively balancing risk management with end-user requirements.

2007 — 2021
Hewitt Associates & Conduent

1. Scaling Domain Complexity & Core Systems Rigor

Spent over a decade mastering the intricate operational, mathematical, and regulatory logic of massive enterprise platforms (US Retirement, Insurtech, Healthcare IT) for top-tier global accounts like MetLife and Allied Life.

  • The Strategy: Transitioned from designing deep execution test cases to translating ambiguous stakeholder objectives into crisp, buildable engineering frameworks.
  • The Product Legacy: Built a career-defining instinct for system resilience, rigorous agile delivery, and managing complex product backlogs where error margins were zero.
2021 — 2023
R Systems International

2. The Catalyst: Translating Affective Computing into Enterprise MVPs

Built a specialized product discovery and delivery team from scratch. This was the turning point: taking experimental computer vision, biometrics, and facial expression analysis out of isolated labs and scaling them into production-ready Emotional AI analytics systems.

  • The Realization: Shaking out these early MVPs revealed the massive gap between a model working flawlessly in a laboratory environment versus a model surviving real-world user data and latency demands at scale.
  • The Inflection: Managing these advanced AI/ML product cycles sparked a deep necessity to go beyond high-level feature specs. To challenge AI limits effectively, I needed to learn how to wire the pipelines myself.
2023 — 2024
Self-Directed Deep Dive

3. Moving from Managing AI Teams to Building AI Infrastructure

Took a calculated professional break dedicated entirely to hands-on AI tooling and technical architecture. Transitioned theoretical ML knowledge into functional development capabilities.

  • The Prototyping Milestone: Designed and custom-coded a localized multi-LLM orchestration architecture using FastMCP, directly proving how disparate model parameters can safely interface with a host desktop ecosystem.
  • The Payoff: This intensive focus on the command line, API routing, and system parameters immediately unlocked the technical autonomy required to engineer and deploy standalone assets like FitCheck completely solo.
2024 — PRESENT
CPP Assistance Services

4. The Modern Product Leader: Code-Grounded Strategy

Synthesizing 18 years of deeply regulated enterprise domain mastery with hands-on technical AI prototyping capabilities to steer next-generation platform roadmaps alongside banking giants like SBI, ICICI, Tata Capital, and HDFC.

  • The Approach: Eliminating product fiction. Every architectural decision, scoping constraint, or model selection I champion is validated by structural testing I’ve performed firsthand.
  • The Impact: Successfully designed, automated, and shipped major platform modernizations—including a self-service customer app—by maintaining complete fluency across strict regulatory landscapes and modern model frameworks alike.

Let's discuss strategy or look at the code.

Open to senior full-time product tracks, hands-on build sprints, or high-level advisory consulting for teams navigating AI/ML systems.

```