Python AI Engineer: Sherlock

Sherlock is a leading blockchain security provider with a mission to create a secure, open financial system. They offer a unique suite of services, including audit contests, bug bounties, and smart contract coverage, and are the only provider to include a warranty after each audit.

Sherlock believes the blockchain security space is on the cusp of a massive AI-driven transformation and aims to be the leader in this disruption. They are seeking a senior Python developer to take ownership of a core feature within Sherlock’s AI product. Your ultimate goal is to build a vulnerability-detecting AI for the crypto space that can mitigate or remove the need for human security researchers. This role is part of a small, elite team in a fast-moving and intense, high-performance culture.

Responsibilities
Take end-to-end ownership of a core feature or process within Sherlock’s AI product, from design through production.

Monitor production AI logs and results to identify failures, regressions, and opportunities for improvement.

Design and run experiments to improve parts of the AI, from prompts and retrieval to model and pipeline changes.

Build and maintain robust benchmarking to measure effectiveness and catch regressions.

Work closely with Sherlock’s developers and security researchers to stay aligned on goals.

Deep dive into blockchain and smart contract vulnerabilities to design practical solutions.

Stay current with practical advances in AI/LLMs and the Solidity/EVM ecosystem.

Requirements
Required Experience:

Senior-level Python experience with a track record of shipping SaaS or software products to production.

Daily use of LLMs and hands-on experience experimenting with or creating AI products or applications.

A willingness to deep dive into the Solidity language and the broader EVM ecosystem.

Nice to Have:

Experience setting up benchmarking and monitoring for AI systems.

Proficiency in Solidity and familiarity with EVM tooling like Foundry or Hardhat.

Familiarity with prompt engineering, evaluation frameworks, retrieval, and vector databases.