Junior Developer / Founding Engineer - Blockchain, AI, Trading Systems

And the less experience you bring, the more the rest has to show: an eagerness to learn that borders on impatience - gaps closed in days, not months; raw sharpness - first-principles reasoning that holds up in any technical conversation; and AI leverage deep enough that your output already reads like a more senior engineer's.

DeFi and Finance - Deep Passion Required

Full fluency isn't expected at this stage - a real, self-driven pull toward this world is. The kind that shows: you've traded, bridged, been liquidated or sandwiched, read the postmortems, followed the space as it moves, and gone down the rabbit holes because you couldn't leave the mechanism unexplained. We teach the rest fast; we can't teach wanting to know.

You'll stand out with a working understanding of:

• Market structure - how liquidity is actually provided across RFQ, order books and AMMs, and what makes execution good or bad;
• The DeFi protocol landscape - how the major protocol families (DEXs and intents, lending, perps, vaults, stablecoins, oracles) are constructed and where they tend to break;
• Settlement and risk - what actually guarantees a trade, who bears the risk while it settles, and how netting and atomicity trade off;
• Tokenized assets - how real-world value gets issued, priced and redeemed on-chain.

Depth in one of these beats surface across all four. The passion is the requirement.

Things We Can Touch - Required

• You've shipped, in public. Deployed projects, open-source contributions, contracts or programs live on-chain, bots that ran - real users, or at least real uptime.
• You're crypto-native. Hackathon builds, protocol contributions, MEV or DeFi tinkering. You know what an RFQ is because you've traded through one, and you have opinions about slippage because you've paid it.

Engineering

A strong programmer first - required. Solid fundamentals, real fluency in a systems-capable language, sound concurrency instincts, and a hard aptitude bar whatever your stack. Rust is a big plus - otherwise show us you absorb languages fast, because you'll write it here.

• Smart contracts - preferred. Deployed Solidity, Anchor or CosmWasm programs, or serious Foundry work: you know the classic exploit classes - reentrancy, access control, oracle games - and why contracts that hold funds are tested differently: fuzzing and invariants, not just unit tests.
• You've built where the markets pulled you - preferred. An arb or market-making bot, a liquidation keeper, a solver, a strategy backtest, on-chain market analytics - something with a price, a position or a PnL in it, run against live chains or venues long enough to teach you what the docs don't: fees and slippage eating the edge, stale prices getting you picked off, races you lost - and why the edge you thought you had wasn't one. You can explain what it taught you.

AI-Native - Required

Where an exceptional early-career engineer outruns experience; your harness and loops are part of your portfolio:

• Agent-first development, and you can show it - spec-first, plan-mode work; rules files with commit history; skills and hooks born from an agent doing something dumb; fresh-context review of agent diffs - tab-completion and a chat window don't qualify.
• Agent orchestration, and you can tell the story - worktree-parallel sessions, unattended Ralph-style loops with verified completion: what state lived between iterations, what broke, what counted as "done".
• Agent and AI infrastructure - preferred - you've stood up some of the stack yourself: an MCP server, RAG or agent frameworks (LangGraph/LlamaIndex-class), a local model - and one real AI feature with the eval most people skip.