Vadodara-based tech hiring startup Utkrusht AI has raised their seed funding round to scale what’s called the world's first Watch-them-Work assessment platform for technical hiring.
The platform takes a completely different approach to a centuries-old problem most tech and engineering leaders still face – you give candidates coding tests or take-home assignments, get results and resumes that look good on paper, interview them, make a hire, and 3 months later wonder how it went wrong or disappointing.
On top of it, tech hiring still takes 4-8 weeks on average to fill out a role.
Utkrusht's core argument is that every current candidate screening + shortlisting methods today – LeetCode/Hackerrank-style tests, coding challenges, take-home assignments, AI interview tools, etc. – have the same flaw, as none can truly mirror the actual conditions of the job.
None of them show you how a candidate actually thinks and works in real job situations (and which now matters even more in the AI-world).
So, their solution is thoroughly researched, rigorously tested, and pretty straightforward in 3 steps:
- Give ALL candidates real on-the-job 30-45min tasks in live production environments (they’ve built a massive infrastructure for doing this)
- Then, just watch them HOW they execute, deploy, solve tasks
- Finally, evaluate their submission on different parameters – technical execution, technical skills, what decisions they made, how they approached the problem, how they used AI, etc.
For eg, some example tasks can be:
- A Backend candidate gets a broken payment microservice to fix
- A DevOps candidate gets a broken Kubernetes cluster to fix, optimize a docker container, etc.
- A Fullstack candidate gets a slow API to debug and deploy
- An AI engineer candidate needs to improve embeddings in a chatbot
- A Data engineer candidate needs to fix kafka partitioning, optimize a hadoop cluster, etc. and more
The task sessions runs for 30-45 minutes, is fully recorded, and candidates can use any tools they want - including AI, and all of their decisions made are captured so you see their approach.
That last part matters. Most assessment platforms today either ban AI use or can't detect it. Utkrusht tracks exactly how each candidate used AI during the session - where, how often, and whether it actually helped them or just generated code they couldn't explain.
Your dashboard shows you a ranked shortlist of the top 5-10 candidates worth interviewing, with detailed reports showing technical execution, problem-solving approach, judgment calls, and AI usage patterns.
The company has now assessed over 10000+ candidates, and its backtesting validates the shortlisted candidates as consistently the strongest quality hires in tech companies. Customers report a ~70% direct slash reduction in time-to-hire, and usually hire in as less than 14 days.
Why This Matters Now
The timing is not accidental. AI has changed what a developer or engineer actually looks like.
The engineers/developers shipping the most today are not necessarily the fastest coders - they are the ones who can navigate ambiguity, make tradeoffs, and direct AI tools toward the right outcomes.
A HackerRank score and report does not tell you any of that. A 30-min recorded work session does.
Arun Bhimani, VP of Technology at InTech Solutions, put it this way: "From 80 applications, I got to watch how candidates think and work before ever meeting them. The person we hired is now leading a core product initiative."
Husna Kouser, Director of HR at Skills Agency, said her team evaluated at least 15 other assessment tools before switching: "None showed us how someone would actually perform on our team."
The funding will go toward expanding the assessment library and improving product, which currently covers 300+ technical skills across a spectrum of skills – fullstack, backend, DevOps, data engineering, SRE, and AI engineering roles.
Free Trial Now Open
Utkrusht is currently offering free access to engineering and tech teams that want to hire better talent in the coming months.
For teams not actively hiring, the free trial is still worth running. Sending a Task link to a few candidates or your internal teammates costs nothing, and the candidate report alone - which shows recorded sessions, AI usage, and a clear hire/no-hire recommendation is a useful benchmark for what good looks like on your team.
Try it out at utkrusht.ai