Hiring from a small team means guessing whether someone can actually do the job, and one startup just raised money to take that guesswork out of the equation. On July 16, DeweyLearn announced in a press release from PR Newswire that they raised a $5 million Series A funding round. The oversubscribed round, as noted in the press release, was “led by SJF Ventures with participation from Catalysis Capital, Morningside and Owl Ventures, among others.”
If you’ve ever hired someone with an impeccable resume who couldn’t do the job, you already understand the problem this solves. Properly assessing skills takes hours you don’t have, so most small teams shortcut the process and absorb the cost later. This shortcut is becoming increasingly difficult to justify.
Inside the $5 Million Round and Who Backed It
DeweyLearn builds a multimodal AI that reads visual, audio, and contextual information generated during a learning activity. Rather than recording what someone clicked or selected, it evaluates how the work was completed.
As the press release above from PR Newswire states, the company works “in the areas of clinical and healthcare education, higher education, professional education, and K-12.” This range is important because assessment methods typically fail when they move from a practical to a knowledge-based domain.
A deployment shows scale. At the Auguste Escoffier School of Culinary Arts, the system made it possible to score more than 20,000 proposals submitted by students, give instructors an initial evaluation and return feedback to students more quickly than before. Arrun Kapoor of SJF Ventures framed the call around better learning outcomes and career access, highlighting the founders’ combined experience in industry and AI.
Why credentials stopped telling you much
Degrees and certificates have always been substitutes, and that substitute is weakening. Skills evolve faster than curriculum updates, so a degree earned three years ago describes a snapshot rather than current ability.
Employers feel it directly. Robert Half’s research found that 62% of hiring managers said skills gaps had worsened in the previous twelve monthswhile only 6% say they have improved.
The frustrating thing is that spending hasn’t fixed the problem. Companies are maintaining their financing programs while employees still feel stuck, which explains so much tension. workplace training which produces certificates instead of capabilities.
| Answer | Share of hiring managers |
|---|---|
| Skills gaps have worsened over the past year | 62% |
| Skills gaps have improved | 6% |
Test skills without building a complete process
You don’t need enterprise software to do better. You need a realistic task that resembles real work, graded based on criteria you noted before anyone applied.
Keep it short and pay if it takes real time. A ninety-minute exercise takes you more than three interviews, and compensating candidates keeps the right people from abandoning your process.
However, write down the grading criteria first. Deciding what good looks like after seeing the job invites you to rationalize a candidate you already liked, which is how bias enters into a process that seemed objective.
This applies well beyond engineering roles. The founders are trying to increase sales often discover that a short mock call predicts performance far better than a careful interview ever did.
Which feedback changes faster in a small team
Speed is the underrated advantage. Escoffier’s example is about review volume, but the real payoff is that feedback comes while the work is still fresh in someone’s mind.
Small teams can copy this without any technology. Review work weekly rather than quarterly and keep feedback specific to a task rather than a personality. People improve when faced with clear examples and linger when faced with vague impressions.
It also changes who you can hire. When you can directly verify your skills, you stop filtering by ancestry and start finding competent people that others have overlooked, which is the number of competent people. Gen Z entrepreneurs ended up building outside the traditional paths.
Hiring and Skills Questions Small Teams Face
Are skills tests fair to candidates? They can be fairer than interviews, as long as the task reflects real work and everyone is graded the same. Problems arise when the exercise tests something that the job never requires.
How long should an assessment take? Less than two hours for most roles. Longer exercises filter free time rather than abilities, quietly excluding people with caregiving responsibilities or current employment.
Should AI score results? First pass, potentially. It is always up to the person to make the decision, since the tool narrows the field while the judgment on suitability remains yours.
Hiring won’t stop being difficult. But replacing an interview with a realistic task is a small change that saves you the much more painful work of fixing a bad hire six months later.
