Open to opportunities

I build intelligent systems
from zero to one.

Staff AI/ML Engineer and startup co-founder. I ship production AI platforms — from multi-agent orchestration over 562K+ clinical trials to real-time trading engines — in Python, Rust, and TypeScript. Columbia Business School. 18+ years in tech.

Python Rust TypeScript PyTorch LLMs / Agents RAG / Embeddings MLOps AWS / GCP React Native Kafka / Airflow
Ish Prasad

Selected Projects

🧬
Production / Healthcare AI

Exmplr

Production multi-agent AI platform for healthcare. Orchestrates 163 specialized tools analyzing 562K+ clinical trials and 75.7M data points. RAG over 19.8M vectors with sub-second semantic search. Multi-provider LLM orchestration (Bedrock Claude, Groq, llama.cpp) with automatic fallback, cost optimization, and quality monitoring via Langfuse.

163 AI Tools 562K+ Clinical Trials 19.8M Vector RAG Multi-Agent Orchestration 90%+ Entity Normalization 99.9% Uptime
Python PyTorch LanceDB AWS Bedrock Airflow Kafka Terraform Cloudflare Workers
📈
AI / Quantitative

QuantoNovo

Production-grade quantitative trading agent implementing Bayesian portfolio optimization with Black-Litterman models, multi-strategy signal generation, and Kelly criterion sizing.

Bayesian Optimization XGBoost / LightGBM Black-Litterman Risk Management 20-Week Roadmap
Python NautilusTrader CVXPY Rust Core
🤖
AI Infrastructure

Minion

High-performance agent framework supporting multi-provider LLM execution, code agents, tool search, MCP integration, and auto-compact context management with streaming.

Multi-Provider LLM Code Execution MCP Protocol Skills Framework Benchmarked
Python OpenAI Anthropic Azure Bedrock
Systems / ML

FlowPulse

Real-time order flow analysis and 0DTE options trading engine in Rust. Processes live market data via WebSocket with Black-Scholes pricing and ML data recording pipeline.

Real-Time Streaming Options Pricing Footprint Candles ML Pipeline Risk Engine
Rust Polygon.io Tradier API WebSocket
🚀
Full-Stack Platform

Leapfrog

Comprehensive Data/AI Engineer career progression platform with interactive SQL/Python labs, AI-powered mock interviews, video curriculum, and cohort management. 28 DB models, 76+ pages.

Interactive Labs AI Mock Interviews Video Platform Cohort System WIOA Compliant
Next.js 14 FastAPI PostgreSQL pgvector Redis Stripe
🕵️
AI Agent

Daemon

Autonomous AI-powered job application agent with browser automation. Supports Full Auto, Semi-Auto, and Assisted modes with intelligent job matching and AI-generated cover letters.

Browser Automation Multi-Mode Agent AI Cover Letters Job Matching Multi-Profile
Electron React Node.js OpenAI Anthropic
🔗
Protocol / P2P

Maaya

Decentralized peer-to-peer agent communication protocol eliminating platform intermediaries. 16 Rust crates, 763 tests, shipping iOS app with native SwiftUI, live Cloudflare relay, on-device encrypted intelligence, and USDC payments on Base.

16 Rust Crates 763 Tests On-Device AI E2E Encryption iOS SwiftUI App USDC Escrow
Rust SwiftUI QUIC Ed25519 Cloudflare Workers SQLCipher

Background

Staff-level AI/ML engineer and startup co-founder with 18+ years building production systems across finance, healthcare, and developer tools. At Exmplr, I designed a multi-agent AI platform orchestrating 163 specialized tools over 562K+ clinical trials, with RAG at 19.8M vectors and multi-provider LLM orchestration.

Prior: 6 years at Citigroup architecting event-driven data pipelines and microservices for financial platforms handling millions of daily transactions. Before that, 6 years at Bank of America building trading and risk management systems. Started my career leading a team of 20 engineers at Microsoft India.

Columbia Business School. Deep across the full ML stack: PyTorch, TensorFlow, Hugging Face for model dev; Airflow, Kafka, Spark for pipelines; MLflow and W&B for experiment tracking. Currently seeking Staff/Founding AI/ML Engineer roles.

18+
Years in tech
562K+
Clinical trials indexed
19.8M
RAG vectors deployed
0→1
Builder mindset

Let's Build Something

Open to Staff/Founding AI/ML Engineer roles.