How TypeSafe came to Eventum for one Senior Product Engineer → and hired all three finalists

On September 15, 2026, TypeSafe launched Jev in early access: AI built to make fast, structured decisions that software can act on directly.

Developers give Jev context and specific questions: Which workflow should handle this request? How relevant is this document? Does this job match our criteria? Jev returns choices, scores, and probabilities within a predefined schema, ready for the application to use. Calibrated probabilities help developers decide when to automate a step and when to send it for further review.

We had already seen the potential firsthand. As an early design partner, Eventum put TypeSafe into production across more than 154,000 job posts. Our benchmark found 104× lower screening costs and approximately 22× lower latency than GPT-5.6 Terra. Read the implementation and results in How Eventum cut LLM screening costs 104x with TypeSafe.

Our partnership also extended to building TypeSafe’s own team.

The challenge

TypeSafe came to Eventum looking for one Senior Product Engineer who could turn its API into compelling demos, useful prototypes, and promising product directions.

The role required technical strength and product instinct: someone who could identify valuable use cases, build quickly, and work independently through ambiguity. They also needed the judgment to recognize what was good enough for a prototype and what would need to change before production.

Our evaluation needed to show those abilities in action.

A hiring process built around the work

We combined targeted outreach with our existing talent network, identifying a strong shortlist in less than a week. Candidates then moved through three focused stages:

1. Alignment conversation. We assessed relevant experience, communication, startup readiness, and fit with the opportunity.
2. Live prototype-building session. Candidates tackled an intentionally ambiguous problem related to the role. They discussed their approach, built a solution, and defended their technical and product decisions, including trade-offs, likely failure points, and scaling considerations.
3. TypeSafe’s cultural round. Our scorecards and notes gave TypeSafe a clear view of each finalist’s strengths before its final conversation.

The live session was the core of the process. The strongest candidates used AI tools effectively, explained their decisions clearly, and were candid about their prototypes’ limitations. They demonstrated the combination of speed, independence, and engineering judgment the role demanded.

The outcome

Within two weeks of kickoff, Eventum had screened roughly 30 candidates and delivered three strong finalists.

TypeSafe had planned to make one hire. After meeting the finalists, it extended offers to all three. All three accepted.

By building the evaluation around TypeSafe’s actual work, Eventum gave the team concrete evidence of each candidate’s abilities—and the confidence to expand a search for one engineer into three hires.