Work

The four MCP servers below are combined ~10,700 lines of code with ~10,000 cumulative NPM downloads. Cited by Pulse and Lobe Hub — two major MCP registries. Two pull requests currently in code review at Anthropic’s official modelcontextprotocol/servers.

Avarieux Inc.

Founder & CEO

Avarieux is a multi-source AI research platform for self-directed investors and registered investment advisors. Every numeric claim is audited against its underlying source before delivery — unverifiable claims are flagged, never silently passed. Every analysis is archived as a permanent, timestamped, citable URL. The system operates under §202(a)(11)(D) of the Investment Advisers Act of 1940 — the publisher exclusion — structurally non-advisory by design. Delaware C-corp, incorporated via Stripe Atlas May 7, 2026.

financial-hub-mcp

Open-source MCP server

A TypeScript MCP server that exposes financial data tools to AI agents — market data, earnings, macro indicators, and more. The highest-download of the four MCP servers, cited by Pulse (pulsemcp.com) and Lobe Hub (lobehub.com).

~3,700 TypeScript Lines of code6,000+ NPM downloads

global-sentinel-mcp

Open-source MCP server

A Python MCP server providing global intelligence data tools to AI agents — geopolitical signals, country-level data, and cross-border information feeds.

~2,200 Python Lines of code

live-audio-intelligence-mcp

Open-source MCP server

A Python MCP server that surfaces live audio processing and transcription capabilities to AI agents — real-time speech analysis, audio classification, and streaming intelligence.

~2,300 Python Lines of code

stealth-agent-browser-mcp

Open-source MCP server

A TypeScript MCP server that gives AI agents browser automation capabilities with stealth evasion — Playwright-based, designed for agents that need to read the live web without triggering bot detection.

~2,500 TypeScript Lines of code

Papex

Founding Engineer

Papex is a NYC fintech building receipt intelligence systems for Indian freelancers — AI-powered financial record-keeping and expense management. Joined as a Founding Engineer, working on the data pipeline and core product infrastructure.

Audio Based Facial Expression Generation On AR Applications

Peer-reviewed publication, IEEE Xplore

Published at the 2023 14th International Conference on Computing Communication and Networking Technologies (ICCCNT), Delhi, India. Third of five authors. Explores real-time facial expression synthesis driven by audio signals in augmented reality contexts.