Technical Screening

A thousand applicants.
The fifty you review.

Candidates solve realistic challenges in a full development environment. Our reviewers evaluate how they solved the problem, and you receive reports detailing what made the best stand out.

See how it works

Sarah Chen

Backend Engineer · 47 min
✓ Recommended
Approach
0:00 Opened api/orders.py
3:12 Checked Grafana for 500s
8:45 Ran EXPLAIN ANALYZE
15:30 Added index, deployed fix
Reviewer Note

"Checked observability before touching code. Validated fix with real traffic. Methodical throughout."

MK
Mike K. · Vettr Reviewer
Interview ✓

How it works

Realistic environment. Objective criteria. Expert review.

01

A real environment

Candidates work in a full development setup—IDE, database, monitoring, terminal. They can deploy their changes and see the results. We capture everything they do.

orders.py
schema.sql
grafana
▾ api
  orders.py
  users.py
▾ db
  schema.sql
  migrations
23 def get_orders(customer_id):
24 # TODO: this query is slow
25 query = """
26 SELECT * FROM orders
27 WHERE customer_id = %s
28 """
~/project $ python -m pytest tests/ -v
02

Objective filters

Before anyone reviews, automated checks verify the solution works. Tests pass, endpoints respond, queries run fast. Only candidates who solved the problem move forward.

Automated Checks All Passed
Test suite ✓ 14/14 passed
API endpoint ✓ 200 OK
Query time 127ms
✓ test_orders_list passed
✓ test_orders_filter passed
✓ test_orders_pagination passed
03

You get reports

We run the same challenge through common LLMs to establish a baseline. Our reviewers compare each candidate's approach against it—seeing what humans add beyond what tools provide. You receive reports on the candidates worth interviewing.

LLM Baseline

Claude 3.5 · 8 min
Baseline
Approach
0:00 Opened orders.py
2:15 Added index directly
6:40 Ran tests, deployed
Notes

Went straight to fix. No investigation of warnings.

Missed: query now returns empty for customer_id 4421

Sarah Chen

Backend Engineer · 47 min
✓ Recommended
Approach
0:00 Opened orders.py
3:12 Checked Grafana for 500s
8:45 Ran EXPLAIN ANALYZE
15:30 Added index, deployed
22:18 Verified with live traffic
Reviewer Note

"Investigated before fixing. Caught the edge case with customer 4421. Solid engineering judgment."

MK
Mike K. · Vettr Reviewer
Interview ✓

Our scenarios or yours

Start immediately with our library, or we'll build challenges from your systems.

Scenario library

Pre-built challenges covering realistic problems—API debugging, slow queries, deployment issues. Ready to use today.

Backend, frontend, full-stack Multiple difficulty levels Calibrated across thousands of sessions

Custom scenarios

We work with your team to build challenges based on real incidents. Your stack, your problems, fully isolated.

Your actual tech stack Tests the exact skills you need We handle all infrastructure

Screen for real engineering

Realistic challenges. Objective filters. Expert review. Get reports on the candidates worth interviewing.