AI Evaluation Protocol Engineer (AI Systems + Blockchain)

The product is not a research memo and not a generic chatbot. It is a working system where operators stake compute, run model evaluation, publish scored results, and settle fees / yield on-chain. We need one developer who can own both sides of that loop: AI — the evaluation methodology, scoring, and job pipeline Blockchain — wallets, staking, settlement, and on-chain proof of those results If you only do ML, the scores never become a market. If you only do Web3, the tokens have nothing real behind them. This role exists because those two pieces have to be designed together. Current AI evaluation is closed, inconsistent, and hard to verify. Different models are scored with different methods, results are easy to fake, and there is no shared ledger for who ran the job, what score was produced, or who gets paid. Stake NXR into evaluation / compute pools Run structured evaluation jobs (benchmarks, scoring, comparison) Publish results as tradable evaluation output / licenses Settle evaluation fees, staking yield, and royalties on-chain We already have a React + Express product shell (marketplace, evaluation studio, staking/trading UI, AI image preview). We do not yet have the real evaluation engine or the on-chain settlement layer. That is the work. What you will do Review existing AI evaluation standards and identify gaps we must cover Design a practical evaluation methodology that works across model types (LLM, vision, agents) Define a scoring framework, result schema, and reporting structure Design how an evaluation result is attested and written on-chain (who ran it, inputs, score, timestamp) Specify the blockchain flows this product depends on: wallet connect, staking pools, evaluation listing, settlement, and claim Turn the above into an actionable build plan we can implement against the existing codebase Must-have experience