Engineering for bot

Hi, I’m looking for an experienced Python/Solana blockchain engineer to help accelerate development of an existing automated cryptocurrency trading research system called MemeQuant Pro. This is not a project starting from scratch. We’ve already spent several months developing the system, and we have an established Python codebase, private GitHub repository, automated testing infrastructure and substantial historical testing evidence. My budget I have a one-off total project budget of $300–$400 USD, not a recurring monthly budget. I’m looking to allocate approximately $200–$260 USD as a fixed-price engineering contract, keeping the remaining budget for API/provider costs and testing. I understand this isn’t enough to rebuild an entire trading platform. I’m specifically looking for someone who can use our existing code and infrastructure to identify and resolve the most important engineering blockers as efficiently as possible. 1. What MemeQuant Pro does MemeQuant Pro is a multi-agent automated cryptocurrency trading research system focused on identifying potentially profitable opportunities in newly launched and emerging meme coins. The system currently includes six specialised agents: * EPSTEIN: Identifies newly launched tokens and potential opportunities. * DIDDY: Performs security, risk and safety checks, with absolute authority to reject unsafe trades. * VILUNÈ: Analyses momentum, market activity and token behaviour over time. * MIA: Combines the agents’ evidence to determine whether a trade qualifies. * MYLO: Investigates potential token revivals and renewed market attention. * KIM: Investigates on-chain wallet behaviour, developers, insiders, whales and suspicious activity. Currently research/detection-only. The broader objective is to have these agents work together in a fully automated pipeline that discovers opportunities, analyses them, rejects unsafe setups and eventually executes trades when a statistically validated strategy is identified. For now, we’re strictly focused on paper trading and proving that the system has a genuine, repeatable trading edge. We are not authorising real-money trading. Our currently active research/execution paths are Solana and Robinhood. Ethereum and BNB remain disabled. 2. Our existing engineering setup We have: * An established Python backend. * A private GitHub repository with extensive development history. * GitHub Actions running on a self-hosted Windows machine using Python 3.12. * Existing integrations with cryptocurrency market-data and blockchain RPC providers, including Helius and DexScreener. * Automated testing scripts and extensive historical evidence. * Established safety checks, signal qualification logic, order/quote validation and shutdown requirements. * A hypothetical $10 position size for paper-trading experiments. We’re already using AI-assisted software development, including ChatGPT/Codex, so you’d be working alongside an existing engineering workflow rather than creating everything manually. 3. The problems we’re currently facing Our biggest challenge is getting the various components to operate reliably together in a genuine end-to-end testing environment. Some of the issues we’ve been investigating include: A. API consumption and data acquisition One genuine provider research run exhausted its allocated 130 DexScreener requests after approximately 208 seconds, before reaching later signal-validation checkpoints at 300 and 360 seconds. We need to understand the actual provider usage, identify legitimate efficiency improvements and establish whether the required data can be obtained reliably within a controlled budget. B. Agent integration and signal timing Our agents have individually undergone considerable testing, but we still need to prove that the original pipeline can reliably process a naturally discovered opportunity through qualification, safety validation, momentum analysis, MIA approval and fresh executable quote validation. We’ve also investigated whether signal expiry prevents some later-stage confirmations. We need this understood properly rather than simply extending deadlines or relaxing safety rules. C. End-to-end paper execution Our immediate goal is to complete a genuine, fully instrumented 15-minute paper-trading session. That includes accurate simulated entries/exits, transaction fees, realistic execution assumptions, safety checks, complete logs and reproducible results. A separate real-provider preflight must qualify before any paper session is authorised. 4. What I’m looking for you to do My priority is not adding new features or redesigning the trading strategy. I’d like you to: 1. Review our existing repository and most recent test evidence. 2. Identify the most important technical blocker preventing the next authentic paper run. 3. Explain the root cause and recommend the smallest reliable fix. 4. Implement the agreed correction within the existing architecture. 5. Run automated tests through our Windows GitHub Actions environment. 6. Provide evidence that the correction works, including logs and test results. 7. Identify any remaining blockers and document exactly what’s required for a safe end-to-end paper run. We have an upcoming offline readiness test planned, so I’d prefer that its results are reviewed first to avoid duplicating previous investigations. 5. Important requirements * Preserve the original trading logic unless I explicitly authorise changes. * Do not weaken or bypass the DIDDY safety veto, signal freshness checks or execution safeguards. * Do not place real trades or connect to wallets capable of spending funds. * Do not purchase paid services or initiate chargeable provider tests without approval. * Work through GitHub with reviewable, documented changes. * Keep sensitive credentials and private research data protected. * Prioritise getting a working, measurable system over unnecessary engineering complexity. 6. Budget and proposed milestones I’d prefer a fixed-price arrangement structured approximately as: Milestone 1 — $60: Technical review, confirmed diagnosis, reproduction of the blocker and a specific proposed solution. Milestone 2 — Up to $200: Implementation, testing and delivery of the agreed fix, including GitHub evidence and documentation. The exact second milestone would depend on what we discover during the initial review. If the full problem cannot realistically be fixed within this budget, I’d rather you explain what can be completed reliably than promise an unrealistic outcome. 7. Questions before we proceed Could you please tell me: * Have you previously developed or debugged Solana trading bots, especially systems involving Pump.fun/PumpSwap, RPC APIs and real-time market data? * Are you comfortable working with a large existing Python codebase rather than building from scratch? * Can you work with GitHub Actions and a self-hosted Windows testing environment? * Would you be comfortable using AI-assisted coding tools where useful to accelerate the work? * What do you realistically believe you could deliver within the proposed $200–$260 engineering budget? * Could you share examples of similar projects you’ve completed and your availability? My longer-term objective is to get MemeQuant Pro operating autonomously, but my immediate priority is making the existing system reliably testable and obtaining genuine paper-trading evidence. If this sounds like a suitable project, I’d be happy to discuss providing controlled access to the repository and the latest test evidence. Thanks.