What they used before
Mix of online job boards + some AI resume-filtering tools
Key Challenges
SkillsAgency's senior developers were conducting ~5 interviews every week with candidates who weren't up to the mark, burning valuable engineering time. Their existing assessment tool (Incruiter) used easily Googleable questions instead of scenario-based tasks, making it ineffective at filtering for real skill. As a agency hiring talent at scale, manual resume screening was impossible—they needed systematic filtering before interviews.
SkillsAgency hires several developers throughout the year for their clients. Their senior developers were drowning in interviewing candidates who weren't up to the mark, while their previous assessment tools weren't shortlisting effectively. They needed something that mirrored actual job tasks and evaluates candidates on that.
Key Outcomes
SkillsAgency eliminated all resume screening time by creating assessment links in 5 minutes and sending them to all applicants from LinkedIn and Indeed. For their first role, 215 candidates took assessments, the platform recommended 8, they interviewed 4, and hired 1 strong developer in 10 days (with only 5 hours invested). They now follow the same playbook for all technical roles, hiring exclusively through Utkrusht's assessment-first approach.
What They Were Doing Before
SkillsAgency was using other popular AI tool in the market, a competing assessment platform. On paper, it should have solved their screening and shortlisting problem. In practice, it didn't.
The assessments were built around multiple-choice questions and simple coding challenges that candidates could easily Google. They weren't testing real skill—they were testing who could search fastest.
Example questions:
"What's the difference between
var,let, andconstin JavaScript?" (Google in 10 seconds)"Which HTTP status code indicates a server error?" (Everyone knows this or can look it up)
"Write a function to reverse a string." (Copy-paste from Stack Overflow)
These questions don't reveal whether someone can actually build software. They reveal whether someone can memorize syntax or use a search engine.
The result? Candidates passed those assessments, then failed SkillsAgency's technical interviews. The tool wasn't filtering effectively—it was just adding an extra step to a broken process.
Manual Screening on Top of Broken Assessments: Even with some tools and software in place, SkillsAgency's team still had to:
Screen resumes manually
Conduct phone screens
Schedule technical interviews with senior developers
Watch candidates fail despite passing coding tests
Their senior developers—50-60 strong team—were doing 2-3 interviews every week, and most candidates weren't up to the mark.
This is the nightmare scenario for an agency: you're supposed to be supplying quality talent efficiently, but instead your own team is stuck screening candidates manually because your assessment tool doesn't work.
Challenges & Pain Points
SkillsAgency faced 3 main problems that their current tools couldn't solve:
1. Senior Developer Time Wasted on Bad Interviews
When you have 50-60 senior developers, their time is your most valuable asset. Every hour they spend interviewing unqualified candidates is an hour not spent:
Writing code for client projects
Architecting solutions
Mentoring junior team members
Delivering value
And SkillsAgency's senior devs were doing 5 interviews per week—most of which ended with "not up to the mark."
At 1-2 hours per interview, that's:
2-6 hours per developer per week on failed interviews
100-300+ hours per week across the team
400-1200+ hours per month burned on candidates who couldn't pass basic technical screens
2. Their In-Built Team Questions Were Too Easy to Game
The fundamental problem with their current methods was simple: the questions tested memorization and Google skills, not real engineering ability.
Candidates would:
Google answers during the assessment
Memorize common responses to predictable questions
Pass the test without understanding fundamentals
Fail the actual technical interview when asked to solve real problems
The assessment layer was supposed to eliminate bad interviews. Instead, it was just adding overhead without improving candidate quality.
3. No Scenario-Based or On-the-Job Tasks
Real engineering work isn't multiple-choice questions. It's:
Debugging production issues under pressure
Building features that handle edge cases
Refactoring messy code to be maintainable
Making architectural trade-offs between speed, cost, and reliability
SkillsAgency needed assessments that mirrored real work—scenario-based problems, on-the-job tasks, situations where candidates had to think and execute, not just recall and regurgitate.
How Utkrusht's Assessment Platform Helped
SkillsAgency didn't just switch assessment tools—they completely changed how they screen candidates. The new process eliminated manual resume review, reduced interview load by ~90%, and gave them confidence that recommended candidates could actually code.
The Key Difference:
1. Scenario-Based Questions That Test Real Engineering Ability
Utkrusht's assessments test whether candidates can solve real problems under real constraints.
Instead of: "What's the difference between == and ===?" (who cares?)
Utkrusht asks: "This React component has a state management bug causing incorrect data to display. Debug it, fix it, and explain how you'd prevent this in production."
This reveals:
✅ Can they debug real issues, or just answer trivia?
✅ Do they understand fundamentals, or just frameworks?
✅ Can they think through edge cases and production scenarios?
✅ How do they approach problems when there's no Stack Overflow answer to copy?
For SkillsAgency, this was the breakthrough. Finally, an assessment that actually predicted interview performance.
2. 0 Time Spent on Resume Screening
The old process:
Post role on LinkedIn/Indeed
Receive 200+ applications
Manually screen resumes (hours of work)
Send promising candidates to Incruiter
Watch them pass Incruiter but fail technical interviews
Start over
The new process:
Create position in Utkrusht (5 minutes)
Get assessment link
Send link to ALL applicants from LinkedIn/Indeed (no resume screening)
Dashboard shows "recommended" vs "not recommended" automatically
Interview only recommended candidates
Hire with confidence
That's it. No manual screening. No wasted time on resumes. No guessing who might be good.
Just: assessment → dashboard → interview the top performers.
3. Dashboard That Shows "Recommended" vs "Not Recommended"
After candidates complete assessments, SkillsAgency's dashboard shows clear recommendations:
✅ Recommended: Candidates who demonstrated strong fundamentals and real execution ability
❌ Not Recommended: Candidates who couldn't pass scenario-based challenges
This eliminated all subjective decision-making. No more "this resume looks good, let's try them." No more "they seem promising, let's give them a shot."
Just objective data: this person can code (recommended) or this person can't (not recommended).
For an agency processing hundreds of candidates, this kind of systematic filtering is transformational.
4. On-the-Job Task watch-them-work tasks, Not Q&A Tests
Utkrusht's assessments don't ask candidates to recite definitions. They ask candidates to complete realistic tasks that mirror actual engineering work:
Debug a broken API endpoint
Optimize a slow database query
Refactor messy code for maintainability
Build a feature that handles edge cases correctly
Fix security vulnerabilities in production code
These tasks reveal whether someone can actually do the job—not whether they studied for the test.
When SkillsAgency's CTOs reviewed Utkrusht's questions, their reaction was immediate: "These questions are good. This tests what we care about."
That validation from technical leadership meant SkillsAgency could trust the platform's recommendations without second-guessing.
The Process Now:
SkillsAgency posts roles on LinkedIn and Indeed (same as before)
They create a position in Utkrusht in 5 minutes and get an assessment link
Every applicant receives the assessment link—no resume screening, no manual filtering
Candidates complete scenario-based assessments testing real engineering ability
SkillsAgency's dashboard shows recommended candidates automatically
They interview only recommended candidates—typically 4-8 out of 200+ applicants
Result: Strong hires in 10 days with minimal time invested
"We don't spend any time screening resumes anymore. We create a position in 5 minutes, get the assessment link, send it to everyone who applied on LinkedIn or Indeed, and our dashboard shows us who's recommended and who's not. It's that simple. No more guessing, no more wasted interviews—just data telling us who can actually code."
- Husna Kouser, Director, HR
Key Outcomes and Results
SkillsAgency's first role using Utkrusht set the template for how they now hire: massive funnel at the top, ruthless filtering in the middle, minimal time invested, strong hire at the end.
The Typical Funnel Now: ~200 → 8 → 4 → 1
~200 candidates take the assessment (all LinkedIn/Indeed applicants—zero manual screening)
8 candidates recommended by the platform (top 3.7%)
4 candidates interviewed by SkillsAgency's team
1 strong developer hired in 10 days
That's a 99.5% reduction in interview load. Instead of interviewing 215 people (or even 50), they interviewed 4—and all 4 were qualified.
Time Invested: 5 Hours Total
New process (with Utkrusht):
Create position and send link: 5 minutes
Review dashboard and select candidates to interview: 30 minutes
Conduct 4 technical interviews (only with qualified candidates): 4-5 hours
Total: 5 hours to make a hire
That's a 90%+ time savings—and they hired someone stronger than they would have found through the old process.
Hire Quality: Strong Developer, Up to the Mark
The developer hired through this process wasn't just adequate—they were exactly what SkillsAgency needed. The assessment had already validated:
✅ Strong technical fundamentals
✅ Real problem-solving ability
✅ Execution under pressure
✅ Code quality and architectural thinking
No surprises. No "we thought they could do it, but..." situations. Just a strong hire who'd proven their depth before the offer letter.
The Playbook: Now Used for All Technical Hires
SkillsAgency didn't just use Utkrusht once and go back to the old way. They now follow the same playbook for every technical role:
Post on LinkedIn/Indeed
Create position in Utkrusht (5 minutes)
Send assessment link to all applicants
Interview only recommended candidates
Hire with confidence in 10-14 days
They've switched from other tools entirely and are hiring exclusively through Utkrusht now. When they took a trial, saw it worked, they bought 500 assessments in bulk. When those ran low, they renewed.
That's the ultimate validation: they're not testing Utkrusht anymore. They're relying on it.
What Stood Out Most
When we asked SkillsAgency what made them switch from other tools to Utkrusht—and what keeps them using it exclusively—they pointed to 3 things: scenario-based questions that actually test skill, the ease of the workflow, and early results that validated the platform immediately.
Scenario-Based Questions vs Googleable Trivia
Utkrusht tested whether you could solve real engineering problems.
When SkillsAgency's CTOs reviewed Utkrusht's questions, they immediately saw the difference. These weren't generic algorithm puzzles or syntax trivia. These were real scenarios:
Production bugs that needed debugging
Performance issues that needed optimization
Code that needed refactoring for maintainability
Features that needed to handle edge cases
This is what engineering actually looks like. And candidates who could handle these scenarios could handle the job.
Workflow So Simple It Requires Zero Training
The second thing that stood out was how easy Utkrusht was to use—not just for SkillsAgency's team, but for their external hiring consultants.
Create position in 5 minutes → Get link → Send to candidates → Dashboard shows recommendations.
That's it. No training required. No onboarding process. No complicated setup.
SkillsAgency's hiring consultants—who supply candidates—could immediately start using the platform without any hand-holding. This was critical for a staffing agency where speed and scale matter.
Early Results Validated the Platform Immediately
The third thing that sealed the deal was the first role's results: 215 candidates, 8 recommended, 4 interviewed, 1 hired—all strong.
But more importantly: the candidates who passed assessments also passed interviews. The candidates who failed assessments would have failed interviews.
This correlation proved the platform was filtering correctly. It wasn't rejecting good candidates or passing bad ones. It was doing exactly what it was supposed to do: separate qualified engineers from everyone else.
That early validation gave SkillsAgency confidence to go all-in. They bought 500 assessments in bulk, made Utkrusht their exclusive technical screening tool, and never looked back.
Why Skills Agency Chose Utkrusht's Platform Over Others
SkillsAgency wasn't looking for a new tool—they already were using different tools. But it wasn't working, so they made the switch. Here's why:
1. Questions That Actually Test Real Skill
Before their questions were Googleable trivia. Utkrusht's questions were scenario-based problems that mirrored real engineering work.
That single difference changed everything. Finally, assessments that predicted interview performance.
2. Zero Resume Screening Required
SkillsAgency before had to still manually screen resumes before sending candidates to assessments. With Utkrusht, they send the assessment link to everyone and let the platform filter automatically.
That eliminated hours of manual work per role.
3. Dashboard That Shows Clear Recommendations
Other tools they used just gave scores. Utkrusht gave recommendations: "interview this person" or "don't waste your time."
That clarity made decision-making instant. No analysis paralysis. No second-guessing. Just: these 8 people are recommended, interview them.
4. Results That Validated Immediately
The first role proved Utkrusht worked: 215→8→4→1 in 10 days with 5 hours invested.
That kind of efficiency isn't luck. It's a system that works.
"For our first role, 215 candidates took the assessment. The platform recommended 8. We interviewed 4 and hired 1 strong developer in 10 days, spending only 5 hours total. That's the playbook we now follow for every technical hire—and it works every single time. We're hiring exclusively through Utkrusht now."
- Husna Kouser, Director, HR
What's Next
For agencies where developer hiring matters so much, speed and quality are everything, and so assessment-first hiring with scenario-based questions and on-the-job tasks isn't optional—it's the only approach that works.
Skills Agency is following the same method now.
"The two developers we hired through Utkrusht are doing really high-quality work. They didn't just pass our technical bar—they raised it. And we didn't have to waste weeks interviewing 50 people to find them. The assessments filtered for depth upfront, so we only talked to candidates who could actually execute."
- Husna Kouser, Director, HR
Want to hire
the best talent
with proof
of skill?
Shortlist candidates with
strong proof of skill
in just 48 hours
Hiring globally across 3 continents, in less than 12 days, and by using pay-per-usage model (not fixed)
Nov 6, 2025
Hiring pre-vetted engineers directly from 1 platform (without worrying about "is this list of candidates able to ship fast?"
Nov 6, 2025
From getting ~250 candidates to hiring in 10 days, spending only 2 hours
Nov 6, 2025
Cutting founder and engg team's initial screening time and filter 200+ applications in 48 hours
Nov 6, 2025