Quant Python Developer — Build Autonomous Trading Bot for Kalshi Prediction Markets (Sports)
Overview
I’m looking for an experienced quant / algorithmic trading developer to build a fully autonomous trading system for Kalshi sports prediction markets using their API.
This is not a generic bot, AI project, or web app.
This is a trading system with:
• Market data ingestion
• Strategy framework
• Risk engine
• Execution engine
• Logging/monitoring
• Paper trading mode
• Dockerized deployment so I can run it locally
If you have built crypto trading bots, betting models, or market-making systems, this is directly in your wheelhouse.
If you have only done web dev, AI prompts, or dashboards, this is not a fit.
Goal
Build a production-ready v1 system that can:
1. Connect to Kalshi API (REST + WebSockets)
2. Monitor sports markets and order books in real time
3. Ingest external data inputs (news/injuries/odds feeds later)
4. Generate trade decisions via pluggable strategy module
5. Execute orders with proper limit/cancel/replace logic
6. Enforce strict risk management rules
7. Run in paper mode first, then live mode
8. Be fully owned and operated by me on my Mac via Docker
System Components Required
1) Data Layer
• Live market data, order book, positions, fills via Kalshi API
• WebSocket listeners + state management
• Rate-limit aware
2) Strategy Layer (framework, not magic)
• Pluggable module where signals/logic can be inserted
• Accept inputs from external scripts/APIs
• Outputs: probability, entry price, size, rationale
3) Risk Engine (critical)
• Max position per market
• Max exposure per sport/day
• Max daily loss
• Liquidity filter (don’t trade thin books)
• Time-to-start filter (no late bad entries)
• Kill switch
4) Execution Engine
• Limit orders by default
• Cancel/replace logic
• Partial fill handling
• Order tracking and reconciliation
5) Modes
• PAPER (no real trades, logs decisions)
• LIVE (real execution)
6) Logging & Monitoring
• Every decision logged
• Every order logged
• Errors + API issues logged
• Simple dashboard or console output is fine
7) Deployment
• Fully Dockerized
• .env based secrets
• One-command start/stop
• Works on Mac locally
Required Skills
You must have experience with:
• Python
• Trading bots / algo trading / betting models
• Order book mechanics
• WebSockets + REST APIs
• Risk management in trading systems
• Docker
• Logging/observability
Huge plus if you’ve worked on:
• Crypto bots
• Sports betting models
• Market making systems
❌ Not a fit if you are
• A web developer
• An AI/LLM prompt engineer
• Someone who has never built a trading system
• Someone who doesn’t understand slippage, liquidity, and order books
Deliverables
By end of project I must have:
1. Private GitHub repo under my account
2. Dockerized system with docker-compose.yml
3. .env.example template
4. Full runbook (install, run, update, troubleshoot)
5. Risk config file with clear parameters
6. Paper trading mode validated
7. Live mode validated
8. Kill switch demo
9. 60–90 min handoff session where I run it locally
10. IP fully assigned to me
⏱ Timeline
2–3 weeks for v1
Budget
Fixed price preferred. Open to milestones.
To Apply (important)
Please include:
1. Examples of trading bots or similar systems you’ve built
2. Your approach to risk management in trading systems
3. Whether you recommend any specific architecture for this
4. Confirmation you are comfortable Dockerizing and handing off clean ownership
Ownership & Security
• All code in my GitHub
• No hardcoded keys
• Work-for-hire IP transfer
Success Criteria
I can run the bot on my Mac, switch between paper/live, control risk settings, and fully understand how to operate it without you..