Italiano English

WHERE THE NEXT BUSINESS BEGINS

← Back to the magazine
Work and automation

Before the interview, engineers must show how they work

Colare brings technical tests forward in engineering recruitment: simulations rather than résumés alone. It could uncover overlooked skills, but who decides what makes a good solution?

FRONTIERA · 2026-10-31 · 5 MIN · 962 WORDS
Illustration of an engineering candidate examining a mechanical component beside a computer
FRONTIERA MAGAZINE.Before the interview, engineers must show how they work
Original AI-generated illustration for FRONTIERA; people, components and settings are imaginary.

A résumé is no longer enough

Before speaking to the hiring team, show how you would design a part or solve an electrical problem. That is the step Colare wants to bring forward in engineering recruitment. The San Francisco startup publicly launched its platform on October 26: simulated technical assessments to evaluate candidates’ work before the interview. For people who are good at doing things but less good at presenting themselves, it could offer a different way in. For employers, it could mean not having to pull designers away from their desks every time.

The founders are Esther Thomas and Nain Abdi. In the launch announcement, Abdi connects the idea to his experience as an electrical engineer at Ford: candidates who looked convincing on paper did not always stand up to technical scrutiny. Colare starts from that gap, without demonstrating that every résumé has now become useless. Previous experience can say a great deal; the question is how relevant it is to the work that needs doing now.

Thomas describes a second obstacle to Payload: applications polished with AI make it harder to distinguish ability from presentation. That is the diagnosis of someone selling a solution, not an independent measure of candidate quality. The proposed response is to shift attention from what a person claims to how they approach a task.

The job becomes part of the test

Illustration of hands comparing two mechanical components beside a computer
Original AI-generated illustration for FRONTIERA; people, components and settings are imaginary. FRONTIERA · Immagine generata con IA

The process begins with the position to be filled. Colare translates the required skills into exercises tied to the discipline: computer-aided design, printed circuit boards, thermal analysis and structural behaviour. The platform compares the process and the result, then returns a detailed evaluation. Getting the right number should not be enough: the assumptions, checks and trade-offs that lead to a solution should count too.

In her interview with SiliconANGLE, Thomas emphasises those trade-offs: a component can become stiffer by gaining weight, or cheaper while becoming harder to manufacture. There is not always a single answer to compare with a model solution. The company says it uses criteria developed with experienced engineers to assess these choices. Those criteria need examining; they do not automatically guarantee that the system recognises every valid solution.

On its product website, Colare describes browser-based exercises involving schematics, code and three-dimensional models. Candidates can complete them alone or alongside the company’s engineers. The assessment distinguishes individual skills and leaves the final decision to the customer. FRONTIERA has read these descriptions: it has not taken an assessment, observed a hiring process or directly verified how the model works.

Who buys time, who offers expertise

The customer is the hiring company. According to Forbes, Colare developed the service with early customers of different sizes, from chemicals to industrial manufacturing. Its promised commercial value is to reduce the work needed to identify credible candidates. The sources consulted do not disclose a price list, so we cannot compare the cost of the service with the time it might save.

Illustration of two engineers reviewing a design together
Original AI-generated illustration for FRONTIERA; people, components and settings are imaginary. FRONTIERA · Immagine generata con IA

The announcement cites John Slack, chief executive of Rhoic, describing the difficulty of assessing an electrical engineer without that specialisation. This is a concrete need for small teams: they may know what problem they need to solve without already having someone in-house who can judge every detail of the solution. The testimony appears in company material, was not collected by us and does not by itself prove the service works.

For a technical manager, a useful assessment should reveal the questions to ask at interview, not simply rank people. An error can indicate a gap in knowledge, but also an assumption different from the one built into the exercise. That is the distinction to look for in the final report: which steps are observable, and which conclusions are the system’s interpretations. Buying a ranking that cannot be questioned would merely relocate the problem.

The filter weighs on candidates too

The promise to jobseekers is to be judged on what they demonstrate, not just their school or most recent employer. Forbes reports the founders’ intention to uncover capable engineers who shine less in conventional recruitment. We do not, however, have independent data on rejected candidates, subsequent hires or their performance. A more concrete test can be better than a generic interview; that does not make it impartial.

Colare also lists webcam and screen monitoring and measures that flag possible outside assistance from AI. Flagging an anomaly is not the same as proving misconduct. Before using a signal to reject someone, a recruiter should be able to examine it and hear an explanation. The pages reviewed do not allow us to quantify errors in these checks or guarantee how they are handled in every hiring process.

Illustration of a candidate concentrating on a computer-based assessment at home
Original AI-generated illustration for FRONTIERA; people, components and settings are imaginary. FRONTIERA · Immagine generata con IA

The time required also matters. Bringing a test forward can prevent pointless interviews, but shifts some effort onto many people who will not be hired. A reasonable exercise needs to be proportionate and relevant, with permitted tools and duration made clear beforehand. Asking for a limited demonstration of ability is different from asking for work the company can use for free: a distinction candidates should be able to recognise.

The test for those doing the judging

The interesting change, then, is when verification happens: technical competence enters the process earlier, while a résumé might still close a door. For this to work, the exercise must truly represent the position. A company that changes duties or tools without updating its test risks selecting for familiarity with the assessment rather than the ability to do the job.

The result to watch is not how many applications the software can examine. It is whether the people selected work well after joining, whether those rejected receive an understandable judgment, and whether engineers recover time without giving up their role in assessment. The open sources contain no independent verification answering all these questions together. For now, they are conditions to check, not established benefits.

For applicants, one simple consequence matters: being able to explain why they made a choice, even when the score does not reward it. If the test opens up that discussion, it adds information to the résumé. If it closes the discussion before a person can intervene, the new entrance to employment will merely have changed its lock.