Building systematic edge through algorithmic scoring, multi-AI consensus engines, and disciplined options strategies on Indian derivatives markets.
Independent derivatives trader and sole developer of a full-stack NSE options trading suite, built from the ground up over two years. I trade NIFTY, BANKNIFTY, FINNIFTY, MIDCPNIFTY options and MCX commodities through Fyers, Zerodha, and Angel One.
My approach combines a 110-point composite scoring engine with a multi-AI consensus debate system — where Claude, GPT, Gemini, and open-source models argue bull, bear, and risk cases before a trade is taken. Every signal passes through volatility gates, institutional flow analysis, and risk governance before reaching execution.
Based in Bengaluru. Building systematic edge, one module at a time.
Every component developed in-house — from data ingestion and scoring to multi-AI debate, risk governance, and execution gateways.
Every trade candidate is scored across seven weighted dimensions — risk-reward, implied volatility, put-call ratio, delta, market profile, option walls, and skew. Scores are computed from live option chain data and fed into downstream decision layers.
An adversarial debate system where multiple AI models argue bull, bear, and risk cases for each trade. A 5-layer pipeline processes OFS data through specialist debate, score parsing, trade history, and an auto-tweaker that proposes alternatives when the judge says avoid.
A FastAPI-based AI research system with 45 registered tools, ChromaDB RAG backend, and a hash-linked audit chain. Provides live trade advisories with option chain data, ATM strikes, premiums, net credit, and multi-point rationale.
A Sinclair 5-layer volatility framework gates every trade — when realized volatility exceeds implied, the system blocks premium-selling entries automatically. Institutional FII/DII flow signals and a portfolio risk governor add additional layers of protection.
Custom AFL strategies feed into a confluence scoring engine. TrueData Velocity provides the real-time data feed. Strategy Ensemble V7 with blink signals, walk-forward optimization, and a signal bus connecting AFL output to the Python trading pipeline.
Trade setups, system development logs, and observations from trading Indian derivatives markets.
A walkthrough of the 5-layer adversarial debate pipeline — from OFS data ingestion to the final judge verdict.
Why blocking trades when realized vol exceeds implied vol saves more than any entry signal.
Using institutional cash market flow data to inform directional bias in weekly NIFTY options.
Interested in discussing trading systems, algo development, or potential collaboration? Reach out through any of the channels below.