A founding engineer should be assessed against the work your company needs next: turning an ambiguous customer problem into a working product, making technical tradeoffs, and staying accountable after launch. Start by defining that work before choosing a title or searching for candidates.
This guide combines a hiring framework with an anonymized lesson from a Refery search. Fictional exercises are labeled separately; it is not a market compensation report.
A real search lesson: broad ownership is not every specialty
In one Refery engineering engagement, the first wave was slow. A recruiting partner identified that the brief combined product engineering with deeper specialist work overlapping another open seat.
Refery put the tradeoff to the founder: could the product role focus on product delivery while the specialist work sat with the other role? The founder agreed that a more conventional full-stack profile could work. Refery returned the answer to the partner.
For a founding engineer search, the lesson is to distinguish broad ownership from an unbounded collection of specialties. Decide what this person must own personally and what support exists around them.
Read the anonymized case study. It is an engineering-search example, not a claim about the eventual hire's exact title or a measured improvement in hiring speed.
Write the first two outcomes
Replace a long technology wish list with two outcomes that would make the hire successful. For example: “Take our prototype through its first customer deployment” and “Make the next deployment repeatable.” This is an illustrative brief, not a Refery client case study.
For each outcome, state what exists today, the constraints, who makes decisions, and how you will recognize success. A candidate should be able to explain what they would investigate before committing to a plan.
Decide which founding engineer you need
| Your immediate problem | Evidence to seek | Question to ask |
|---|---|---|
| Build a usable product from a prototype | End-to-end ownership and customer feedback | What did you choose not to build, and why? |
| Make a fragile system reliable | Debugging, operational judgment, and observability | Walk through an incident you owned after launch. |
| Deliver an AI feature users can trust | Evaluation design, failure analysis, and product integration | How did you decide the system was good enough to release? |
| Get a product working inside customer environments | Integration depth and customer communication | Which customer constraints changed your implementation? |
If the last row describes most of the job, read the forward deployed engineer hiring guide. The title matters less than the actual work.
Use a scorecard with evidence, not adjectives
Score the same dimensions for every candidate: problem framing, delivery, technical judgment, ownership after launch, and collaboration. Write down the evidence before discussing an overall recommendation.
Use four labels: no evidence yet, partial evidence, meets the requirement, and exceeds the requirement. “No evidence” means you need another question; it does not mean the candidate lacks the skill. Identify essential requirements before meeting candidates so a strong presentation does not quietly replace the hiring bar.
Run a focused interview sequence
- Scope conversation: explain the company's constraints and ask the candidate to identify the unknowns.
- Project deep dive: ask what they personally built, the alternatives they rejected, and what happened after release. Separate team achievements from individual contributions.
- Bounded work discussion: use a fictional or sanitized problem resembling the job. State the time limit and evaluation criteria in advance. Do not ask candidates to perform unpaid production work.
- Working relationship discussion: cover disagreement, decision ownership, feedback, and how the founders will support the role.
- References with consent: ask former collaborators about a specific project and the candidate's responsibilities.
Assign each interviewer a different question to resolve. Repeating the same project conversation adds meetings without necessarily adding evidence.
Make the offer understandable
Explain cash compensation, the equity instrument and grant terms, location expectations, reporting line, and responsibilities separately. Avoid describing speculative equity value as guaranteed cash compensation. Use current, relevant compensation sources and candidates' feedback rather than treating one salary range as universal.
The search should also explain why this person would choose your company. Give them access to the founders, the problem, and a realistic picture of the next stage.
Before you start sourcing
You should be able to share a one-page brief, a bounded interview process, a decision owner, and an agreed budget. Use the recruiting partner checklist if you need outside help.
Hiring with Refery
Refery helps early-stage startups hire engineering, AI, and GTM talent through introductions from founders, operators, investors, and specialist recruiters, with screening for the role. The current service and pricing are 15% to 25% of first-year base salary, paid only when you hire; scope and fee are agreed upfront.
Share your role with Refery, or use the hiring brief template to prepare your search.
If you are choosing how to run the search, the best startup recruiting platforms and networks in 2026 groups the options by model.