Full-Stack Developer (Python + JavaScript, AI-Assisted) — Real-Time Betting Operations Platform
Budget: $8 – $15 AUD
Overview
We run a real-time sports-betting operations platform and are looking for a strong full-stack developer to join ongoing development. It's a live, in-production system — you'll be shipping features and hardening existing ones, not greenfield scaffolding. The platform ingests betting tips/odds from multiple live sources, presents them on a real-time operations dashboard, and includes a browser-automation toolchain that places and tracks bets across major Australian bookmakers.
We build extremely fast with heavy AI assistance. A large amount of this codebase is written with AI coding tools, and we expect you to do the same — use AI to move quickly while still understanding, reviewing, and owning every line you ship. If you're a strong engineer who has learned to drive AI tools to multiply your output, you'll fit right in. If you're slow without them, or blindly paste AI output without review, you won't.
How we work
- AI-first velocity: Claude Code / Cursor / Copilot (or your tool of choice) are part of the daily workflow. We ship multiple iterations a day.
- You own the output: AI writes a lot of it; you're responsible for correctness, edge cases, and not breaking a money-adjacent live system.
- Tight feedback loop, direct contact, quick decisions — no heavy process.
The stack
- Backend: Python (Flask), SQLite, Server-Sent Events (SSE) for real-time updates
- Frontend: HTML/CSS/vanilla JavaScript (single-file dashboard), real-time SSE-driven UI
- Automation: Chrome MV3 extension (service worker + content scripts) that automates interactions with bookmaker websites
- AI in the product: LLM-powered parsing (Anthropic/Claude) for turning messy tip text/images into structured bets
- Integrations: reverse-engineered bookmaker read APIs (live odds/markets/prices), plus Discord/Telegram data scrapers and a live-odds middleware
- Infra: self-hosted, PM2-managed services, deployed via Git pull + restart
What you'll work on
- New features on the operations dashboard (real-time bet feed, placement tracking, reconciliation, reporting)
- Extending the browser-automation engine: new markets/bet types, more bookmakers, reliability hardening (timeouts, retries, guards, state recovery)
- Reverse-engineering and integrating additional bookmakers' public JSON/GraphQL read APIs
- LLM-powered data pipelines: ingesting and normalizing tips/odds from various sources
- Reconciliation and audit logic (matching placements against bookmaker transaction history)
Must-have skills
- Strong Python (Flask or similar) and JavaScript
- Fluent with AI-assisted development (Claude Code / Cursor / Copilot, etc.) — proven ability to ship fast with AI while reviewing and owning the result
- Experience building/maintaining Chrome MV3 extensions (service workers, content scripts, messaging, MV3 lifecycle/limitations)
- Comfortable reverse-engineering web APIs (readins, replicating requests, handling authheaders/UA/anti-bot)
- Solid with real-time systems (SSE/WebSockets) an
- Defensive coding mindset — guards, validation, idempotency, graceful failure (this system handles money-adjacent actions)
Nice to have
- Background or strong knowledge in statistics / probability (we work with odds, EV, value, variance — being able to reason about
the numbers is a real plus)
- Experience integrating LLMs into products (Anthropic/Claude API, prompt design, structured output)
- Browser automation experience (Playwright/Puppet
- Familiarity with sports betting / trading / bookmaker platforms
- Experience with virtualized/SPA scraping and fla
- Node.js, PM2, basic Linux/Windows server ops
Engagement
- Part-time to begin with, flexible hours, hourly at AUD $10–15/hr
- We start everyone with a small paid trial task tnto the ongoing backlog — strong performers getmore hours over time
We run a real-time sports-betting operations platform and are looking for a strong full-stack developer to join ongoing development. It's a live, in-production system — you'll be shipping features and hardening existing ones, not greenfield scaffolding. The platform ingests betting tips/odds from multiple live sources, presents them on a real-time operations dashboard, and includes a browser-automation toolchain that places and tracks bets across major Australian bookmakers.
We build extremely fast with heavy AI assistance. A large amount of this codebase is written with AI coding tools, and we expect you to do the same — use AI to move quickly while still understanding, reviewing, and owning every line you ship. If you're a strong engineer who has learned to drive AI tools to multiply your output, you'll fit right in. If you're slow without them, or blindly paste AI output without review, you won't.
How we work
- AI-first velocity: Claude Code / Cursor / Copilot (or your tool of choice) are part of the daily workflow. We ship multiple iterations a day.
- You own the output: AI writes a lot of it; you're responsible for correctness, edge cases, and not breaking a money-adjacent live system.
- Tight feedback loop, direct contact, quick decisions — no heavy process.
The stack
- Backend: Python (Flask), SQLite, Server-Sent Events (SSE) for real-time updates
- Frontend: HTML/CSS/vanilla JavaScript (single-file dashboard), real-time SSE-driven UI
- Automation: Chrome MV3 extension (service worker + content scripts) that automates interactions with bookmaker websites
- AI in the product: LLM-powered parsing (Anthropic/Claude) for turning messy tip text/images into structured bets
- Integrations: reverse-engineered bookmaker read APIs (live odds/markets/prices), plus Discord/Telegram data scrapers and a live-odds middleware
- Infra: self-hosted, PM2-managed services, deployed via Git pull + restart
What you'll work on
- New features on the operations dashboard (real-time bet feed, placement tracking, reconciliation, reporting)
- Extending the browser-automation engine: new markets/bet types, more bookmakers, reliability hardening (timeouts, retries, guards, state recovery)
- Reverse-engineering and integrating additional bookmakers' public JSON/GraphQL read APIs
- LLM-powered data pipelines: ingesting and normalizing tips/odds from various sources
- Reconciliation and audit logic (matching placements against bookmaker transaction history)
Must-have skills
- Strong Python (Flask or similar) and JavaScript
- Fluent with AI-assisted development (Claude Code / Cursor / Copilot, etc.) — proven ability to ship fast with AI while reviewing and owning the result
- Experience building/maintaining Chrome MV3 extensions (service workers, content scripts, messaging, MV3 lifecycle/limitations)
- Comfortable reverse-engineering web APIs (readins, replicating requests, handling authheaders/UA/anti-bot)
- Solid with real-time systems (SSE/WebSockets) an
- Defensive coding mindset — guards, validation, idempotency, graceful failure (this system handles money-adjacent actions)
Nice to have
- Background or strong knowledge in statistics / probability (we work with odds, EV, value, variance — being able to reason about
the numbers is a real plus)
- Experience integrating LLMs into products (Anthropic/Claude API, prompt design, structured output)
- Browser automation experience (Playwright/Puppet
- Familiarity with sports betting / trading / bookmaker platforms
- Experience with virtualized/SPA scraping and fla
- Node.js, PM2, basic Linux/Windows server ops
Engagement
- Part-time to begin with, flexible hours, hourly at AUD $10–15/hr
- We start everyone with a small paid trial task tnto the ongoing backlog — strong performers getmore hours over time