Senior Full-Stack Engineer — Sports Prediction Market Trading Platform

About the project We operate a live sports prediction-market copy-trading platform that monitors selected trader wallets, classifies their activity, generates betting signals, and executes orders across multiple venues. Our platform currently includes: A central PolyWalletTracker application for managing tracked accounts, monitoring wallet activity, generating signals, and configuring copy rules A Polymarket trading executor Separate Kalshi and 4casters trading executors Multi-venue rule engines covering leagues, market types, sides, position sizing, signal matching, opposing signals, order repricing, partial fills, cancellations, and risk controls Multiple repositories and Railway services rather than a monorepo The system is live and manages meaningful internal trading capital. Reliability, visibility, and safe execution are more important than simply shipping features quickly. We are looking for a senior engineer who can take ownership of the entire platform, stabilize the current architecture, improve operational reliability, and work closely with the founders and existing development team. The role This is a full-time Senior Full-Stack Engineer position with significant backend, infrastructure, and production-operations responsibility. You should be comfortable independently investigating incidents, restoring services, reviewing risky logic, improving architecture, and communicating quickly with the team. Responsibilities You will own and improve the complete platform, including: Maintaining the TypeScript, Node.js, Next.js, React, and Python services Understanding and documenting all repositories and Railway deployments Stabilizing the Polymarket, Kalshi, and 4casters execution systems Maintaining integrations with the Polymarket CLOB API and official Kalshi and 4casters APIs Improving wallet polling, webhook ingestion, trade detection, signal generation, and market matching Improving the copy-rule engine, risk limits, daily budgets, and emergency controls Handling limit orders, cancellations, repricing, partial fills, and resting orders Preventing duplicate, stale, incorrect, or runaway orders Improving position and balance reconciliation across trading venues Separating critical workers from web services where appropriate Improving PostgreSQL schema management and introducing safe migrations Improving Redis usage, background processing, and service coordination Adding automated tests around financially sensitive execution logic Introducing proper logging, monitoring, alerts, and incident reporting Improving deployment safety and recovery procedures Reviewing code written by other developers Helping manage and guide the engineering team Proactively identifying risks and recommending better technical practices Current technical environment The current platform uses a combination of: TypeScript Node.js 20+ Python Next.js 16 with the App Router React 19 Tailwind CSS PostgreSQL Redis Raw SQL Railway Server-Sent Events HTTP polling and webhooks Telegram Bot API Polymarket CLOB client Ethers.js Official Kalshi API Official 4casters API The platform is distributed across multiple repositories and Railway services. Some long-running workers currently run inside web processes, and parts of the trading engine contain large, stateful modules that need careful refactoring. Reliability challenges The main problem we are solving is not a lack of features. It is a lack of dependable technical ownership. The current system can experience different issues on a daily basis. When an engineer becomes unavailable, the team may not immediately know how to diagnose the problem, restore the affected services, or safely confirm that order execution has resumed. The new engineer must help ensure that: The team is never dependent on one unavailable developer Every service has documented recovery procedures Trading failures are detected immediately Failed services can be restored safely Open exchange orders remain visible during deployments or outages A Railway restart or deployment cannot create uncontrolled trading behavior Repeated order placement cannot continue unnoticed Critical execution logic has automated tests and safeguards Required experience Candidates should have: At least 6 years of professional software-engineering experience Strong TypeScript and Node.js experience Strong full-stack experience with React and Next.js Experience operating production financial, trading, betting, exchange, or real-time transaction systems Strong PostgreSQL knowledge Experience with Redis and background workers Experience designing idempotent systems Experience handling retries, duplicate events, race conditions, and partial failures Experience with REST APIs, webhooks, polling, and real-time event systems Strong debugging and production incident-response skills Experience deploying and maintaining multi-service applications Experience protecting private keys, API credentials, and systems capable of moving funds Ability to work independently without needing detailed instructions for every task Ability to review an unfamiliar production system and quickly understand its architecture Good written English and fast communication Railway-specific experience is helpful but not mandatory. Strong Docker, cloud infrastructure, and production deployment experience is more important. Strongly preferred Experience in one or more of the following areas is highly valuable: Sports betting Prediction markets Polymarket Kalshi 4casters Exchange order books Crypto wallet signing CLOB APIs Automated trading Copy trading Market and line matching Moneyline, spread, and totals markets Order repricing Position reconciliation Risk engines and kill switches The ideal candidate understands that a sportsbook-style market, a prediction-market contract, and an exchange order book may represent the same underlying sporting outcome differently. Working schedule and communication The team communicates primarily through Telegram. The engineer must be actively working and available between: 5:00 PM and 8:00 PM Eastern Time During this three-hour window, the engineer should: Be at their computer Monitor Telegram Respond to urgent messages in approximately one minute Be ready to investigate production issues immediately Coordinate actively with the founders and development team Outside this window, the engineer will have flexibility in organizing their work. However, because this is a live trading platform, the role includes responsibility for urgent production incidents and system availability. The role does not require frequent meetings. Fast written communication, ownership, and execution are more important. What success looks like First 30 days Understand all repositories, services, databases, and Railway deployments Document the full signal-to-order architecture Create a clear service recovery and incident-response guide Identify the highest-risk execution and infrastructure issues Add monitoring for critical services and trading failures Ensure the founders can understand whether every executor is operational Add safeguards against duplicate or runaway orders First 90 days Reduce dependence on individual developers Make service recovery predictable and documented Add automated tests around critical trading logic Improve deployment and rollback safety Improve logging, alerting, and reconciliation Refactor the most dangerous stateful execution components Ensure Polymarket, Kalshi, and 4casters behave consistently Independently own production releases and incidents Help establish reliable engineering practices for the rest of the team Who should not apply This role is not suitable for someone who: Regularly disappears during working hours Treats production support as optional Requires detailed instructions for every implementation Has only frontend experience Has never maintained a live production system Is uncomfortable working with financial risk Cannot communicate quickly during incidents Wants to contribute only a few hours per week Relies entirely on AI coding tools without being able to debug the underlying system Using Cursor or other AI development tools is welcome, but the engineer must personally understand, verify, test, and take responsibility for the resulting code.