Enhancement of Python-Based Nifty 50 Analyzer
Budget / Salary₹1,500–12,500
TypeFreelance project
LocationRemote
Posted1 hour ago
I have a working Nifty 50 analyzer (Python/FastAPI, vanilla JS, Kite Connect), live on a Linux VPS. I need an experienced developer to add a professional, prop-desk-style AI analyst on top of it.
Core deliverable: AI analyst engine
OI analysis: option chain OI and change in OI, long/short buildup and unwinding, PCR, max pain, support/resistance from OI walls, and strike-wise OI-inflow heatmap bars in the option chain.
CVD / order flow: cumulative volume delta and delta divergence on Nifty futures, built from Kite tick/depth data (index has no real volume), with session and rolling views.
Technical analysis: multi-timeframe trend, VWAP, key levels, market structure, and volatility regime (IV/ATR).
News integration: a news API feed (TrueData or similar), tagged by impact and sentiment, feeding into the analysis.
Analyst output: a structured trade view in prop-desk format: bias, confluence score, entry, stop, targets, invalidation, and risk-reward, with a plain-English rationale.
Architecture requirements
Numbers must be computed deterministically in code. The LLM only interprets and summarizes them, and it must never invent data.
Structured (JSON) outputs, with every signal logged for later review.
A confidence/no-trade filter, since "no setup" is a valid output.
Also include (my additions)
Backtesting and signal-validation framework with win rate, expectancy, and drawdown.
Paper-trading mode with a trade journal. No live order execution in this scope.
Risk module: position sizing, daily loss limit, kill switch (design only).
Reliable Kite daily token refresh and reconnect handling on the server.
Clean UI panel for the analyst in the existing dashboard.
Documentation and a short handover session.
Required skills: Python, FastAPI, Kite Connect (or similar Indian broker API), options/OI analytics, order flow, LLM integration, Linux VPS (nginx, systemd).
Security and process
No live API keys or root access shared. Work happens on a staging environment with a limited-permission user.
Delivery via GitHub pull requests in milestones; I deploy to production myself.
Please start with a short paid milestone (OI + CVD module) before full scope.
To bid, please send: a similar project or repo, your milestone breakdown, and your timeline. Experience with Indian markets is strongly preferred.
Core deliverable: AI analyst engine
OI analysis: option chain OI and change in OI, long/short buildup and unwinding, PCR, max pain, support/resistance from OI walls, and strike-wise OI-inflow heatmap bars in the option chain.
CVD / order flow: cumulative volume delta and delta divergence on Nifty futures, built from Kite tick/depth data (index has no real volume), with session and rolling views.
Technical analysis: multi-timeframe trend, VWAP, key levels, market structure, and volatility regime (IV/ATR).
News integration: a news API feed (TrueData or similar), tagged by impact and sentiment, feeding into the analysis.
Analyst output: a structured trade view in prop-desk format: bias, confluence score, entry, stop, targets, invalidation, and risk-reward, with a plain-English rationale.
Architecture requirements
Numbers must be computed deterministically in code. The LLM only interprets and summarizes them, and it must never invent data.
Structured (JSON) outputs, with every signal logged for later review.
A confidence/no-trade filter, since "no setup" is a valid output.
Also include (my additions)
Backtesting and signal-validation framework with win rate, expectancy, and drawdown.
Paper-trading mode with a trade journal. No live order execution in this scope.
Risk module: position sizing, daily loss limit, kill switch (design only).
Reliable Kite daily token refresh and reconnect handling on the server.
Clean UI panel for the analyst in the existing dashboard.
Documentation and a short handover session.
Required skills: Python, FastAPI, Kite Connect (or similar Indian broker API), options/OI analytics, order flow, LLM integration, Linux VPS (nginx, systemd).
Security and process
No live API keys or root access shared. Work happens on a staging environment with a limited-permission user.
Delivery via GitHub pull requests in milestones; I deploy to production myself.
Please start with a short paid milestone (OI + CVD module) before full scope.
To bid, please send: a similar project or repo, your milestone breakdown, and your timeline. Experience with Indian markets is strongly preferred.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.