How to choose an R&D partner for your deep tech startup

The wrong engineering partner can cost you a year and a round of funding. The right one can compress your roadmap, give your investors confidence, and let you focus on the company instead of the build.

That's a real range of outcomes. It's worth thinking carefully about who you pick.

This is written for founders of deep tech, hardware, or hardware-software companies who are considering working with an outsourced R&D or engineering firm. The advice applies whether you're hiring for a small piece of work or planning to embed an external team alongside yours.

Why this is harder than hiring a person

When you hire an engineer, you're evaluating one person's skills against one role. When you hire a firm, you're evaluating their judgment, their bench, their process, and their ability to scale their team up and down as your project changes.

You're also evaluating people you may never meet. Most firms staff projects with whoever is available when you sign, not the senior people who showed up in the sales meeting. That's the single biggest gap between expectation and reality.

The good news is that you can tell who's serious within a few conversations if you ask the right questions.

What to actually look for

A few things separate firms worth working with from firms that look good on a website.

Depth in your specific area, not generic "we do hardware" language. Have they shipped something close to what you're trying to ship? Optics, thermal, power electronics, embedded AI, whatever your specific domain is. Ask for examples. Ask which engineers worked on those projects. Ask if those engineers will be on yours.

Senior people who stay on the project, not just on the pitch. Some firms are excellent at sales and average at delivery because the senior engineers move on once the deal closes. Ask who specifically will be doing the work, what their experience is, and whether you'll be talking to them weekly.

A real process for ambiguous work. Deep tech projects almost never run in a straight line. Ask how they handle scope changes, surprises, and decisions that need to be made quickly. A vague answer here is a warning sign.

Credibility with people who matter to you. If you're planning to raise on the back of this work, does your firm's name carry weight with investors and partners in your space? You can check this with two phone calls.

Clear IP and commercial terms. Read the master services agreement before you fall in love with the firm. Make sure IP assignment is unambiguous, that subcontracting is controlled, and that the engagement structure makes sense for what you're trying to do.

Red flags

A few things that should make you slow down.

A pitch that hits every capability without picking one. Firms that "do everything" usually do nothing exceptionally.

No examples they're willing to talk about. Confidentiality is real, but a serious firm has at least a few case studies, references, or anonymized stories they can share.

Junior engineers on the call when senior ones were promised. If the team you meet in the sales process isn't the team that shows up in week one, the project is starting badly.

Resistance to fixed-scope work when the project is well-defined. Some work genuinely needs time-and-materials. A good firm can do both. Insistence on T&M for a clearly bounded project is sometimes a sign of weak estimation.

Pressure to sign quickly. The wrong firm can hurt you more than no firm. Take the time.

Good questions for the first call

A few questions worth asking, in some form, before you sign anything.

Have you done something close to this before? Who specifically worked on it?

Who will be on my project, week to week? What's their background?

How do you handle a surprise that adds a month to the timeline?

What's your average client size, and where do I fit in that range?

What's a project you walked away from, and why?

How do you charge, and what does a typical engagement actually cost end to end?

Their answers tell you a lot. So does the way they answer.

Engagement structures

A quick map of how these relationships are usually structured.

Fixed scope, fixed fee. Best for well-defined projects with stable requirements. Predictable for you. Risky for the firm if scope drifts, so expect tight change-management.

Time and materials. Best for ambiguous or research-heavy work where the path isn't clear. Requires more management from your side.

Embedded team. The firm provides a senior team that works alongside your in-house people for an extended period. Best when you want continuity and the work is ongoing.

Phased engagement. A short, fixed-scope discovery phase followed by a larger engagement once both sides understand the work. This is usually the safest way to start, especially when neither side has worked with the other before.

The honest version

Picking an R&D partner is one of the higher-stakes commercial decisions a deep tech founder makes. Treat it that way. Take three or four conversations. Talk to references. Read the agreement. Negotiate the parts that don't fit your situation.

The firms worth working with will appreciate that you're being careful. The ones that don't are telling you something.