Applied AI Engineer

About Us

Benchmark is the trusted AI platform for investment firms. Today, leading firms use Benchmark to screen deals, extract insights from unstructured data, and automate workflows. Our platform combines cutting-edge AI with deep industry expertise to help investment firms make smarter, faster decisions across their entire deal lifecycle.

What we are looking for

• 3+ years of experience shipping production applications
• Strong Python skills and deep experience with modern ML/AI frameworks
• Genuine interest in LLM's and keeps up to date with current research + model capabilities
• Experience with embedding systems, vector databases and retrieval architectures
• Self-motivated, high ownership and low ego with the ability to work through ambiguity
• Excited to work in-person in our Soho office 3+ days/week
• Bonus
  • Have experience working with LLMs in production
  • Have previous experience working at a start-up

Things you would work on

• Driving user value with LLMs: We are constantly improving our AI systems, which includes:
  • Building and optimizing our core retrieval and processing architecture
  • Testing and deploying new models and approaches
  • Designing, optimizing and testing prompts
  • Building and running comprehensive evals
  • Creating scalable solutions for complex information processing challenges
• Owning our AI infrastructure: This hire will be expected to be the expert on:
  • Architecting and maintaining our core AI infrastructure
  • Driving rapid prototyping of ML/AI solutions
  • Making strategic decisions about model selection and deployment
  • Building sophisticated NLP systems beyond basic LLM integrations
  • Designing advanced information processing pipelines

Our Tech Stack

• Backend: Python, Flask, Postgres
• Frontend: Typescript, React
• Infra: GCP

Why Us

• We believe that in person work matters - we are trying to ship ambitious products with a lean team and the time we spend together in person is a competitive advantage.
• Our technical DNA comes from Square, Google, OpenGov, Carta, etc - everyone builds, is senior and works autonomously.