Back End Developer (Go) — Crypto Trading Systems

Responsibilities

• Take ownership of one of our production trading systems (Go and NodeJS/TypeScript), starting with a defined handover and a runbook — then keep it fast, stable and profitable.
• Hunt down edge cases before the market does: our systems trade real money at high speed, so correctness and observability (Grafana, Prometheus, logs) are the core craft.
• Work AI-first: we develop with AI coding agents (Claude, Codex) daily. Your job is to steer them well and to verify what they produce — which is why fundamentals matter more here than framework familiarity.
• Research and implement improvements to strategy logic together with the PM, and — as your scope grows — contribute to new strategies with a PnL share attached.
• Grow into research. As you master your system — at your pace, not a fixed timeline — an increasing share of your time becomes research: hunting new competitive edges in existing strategies, testing ideas for new ones, digging into niche mechanisms at the finest level of detail. If you only ever want to write code, this drift will feel boring or too hard. If you want to explore, test ideas against a live market, and find edges nobody has published, it is the best part of the job.
• Document what you build so any teammate could run your system: replaceability is an explicit duty here, in both directions — you will inherit well-documented systems and leave them better documented.

Requirements

• Strong programming fundamentals — data structures, concurrency, networking, debugging under pressure. This is the one thing we cannot teach and will test hardest. We can teach you EVM; we cannot teach you fundamentals.
• Go, plus working NodeJS/TypeScript — our performance-critical systems are Go; much of the estate is NodeJS. Strong Go engineers who can operate comfortably across both fit best.
• Comfort owning brownfield production systems — inheriting a live system built by someone else, understanding it deeply, and improving it without breaking it.
• Production ownership mindset — monitoring, logging, incident response, postmortems; willingness to support live systems outside office hours during critical market windows.
• Basic mathematical fluency — algebra and quantitative reasoning; you will be working next to pricing formulas and pool math.
• Full commitment — this role is not compatible with side businesses or parallel commercial projects. We pay for results and share profits precisely because we ask for your full professional focus.

Nice to have

• EVM / DeFi experience (Uniswap-style AMMs, mempool, gas mechanics) — people with it onboard noticeably faster.
• Solidity / smart-contract development.
• Algorithmic trading or market microstructure understanding.
• High-load, low-latency system experience.

Benefits

• Salary negotiable + performance bonuses at bi-annual reviews.
• PnL share on the strategies you work on — eligibility starts when you pass probation, not at an anniversary. The share is computed on strategy profit net of your own fully-loaded cost, under a formula you see in writing — so it pays exactly when your work is genuinely in the money, whether that takes six months or two years. The ladder that widens your share toward the whole book is measured in demonstrations, not months: an exceptional first year climbs it fast.
• Transparent levels and access — written level definitions in observable terms, and a published access ladder (source access in a trading firm is earned with track record; the rungs and criteria are explicit from day one, and none of them is a waiting period).
• A path to running your own book — developers who prove out can originate and run strategies on the firm’s capital and infrastructure, with economics attached. Ambition has somewhere to go without leaving.
• Structured onboarding, wherever you sit — Hanoi (with the dev team) or Saigon (with the PM, in office ~3 days/week). Either way you get formal scheduled mentor sessions — daily in your first two weeks, several times a week through probation, weekly after — plus an async question channel with a committed answer time. Onboarding here runs on schedule and structure, not on hoping someone is free.
• 12 days annual leave; a small, senior, low-meeting team; AI tooling budget — we build with the best agents available.

Interview Process

• 1. Coding assessment — remote, at your own time, without AI assistance or web search. It tests fundamentals, not trivia. (We build with AI daily; the test exists because steering AI well requires exactly the fundamentals it measures.)
• 2. Math evaluation — the math you will actually meet in the work: Uniswap/AMM math, interest and funding calculations, buy/sell arbitrage reasoning. If you enjoy this test, you will probably enjoy the job.
• 3. Technical + fit discussion — face to face. You will walk through your own coding submission, explain your choices, and extend it live. Then: your approach to inheriting a production system, how you work with AI agents, incident stories, and an honest two-way conversation about whether the bargain above is one you actually want.
• Speed commitment: if you pass every round, you will have an offer within one week of taking your first test. We arrange the rounds to make that true, because we respect your time and we know good engineers do not stay on the market.