The archive · Developer & Business Tools · Product decision · 2022
Litebulb (YC W22) bet real-codebase coding interviews would beat LeetCode
YC-backed Litebulb automated coding interviews as real work on real codebases, not LeetCode, launching in March 2022 with companies already paying.
Litebulb
What the business is
SaaS platform for employers: candidates get a Dockerized real codebase, build a feature per specs, submit a pull request, and Litebulb runs tests, load and static analysis, then returns a hiring report.
How it started
Gary Lin, who had worked at 11 companies and been through well over a hundred interview funnels, founded Litebulb after concluding that puzzle screening tested the wrong skill while interviews drained engineer time. The company went through the YC W22 batch, and by early 2022 it was hiring its own engineers through the product and had paying customers before the public launch.
What happened
The Launch HN went up on 2022-03-07 and drew 139 points and 190 comments. The pitch was concrete: interview environments are Dockerized, setup is boiled down to a single make command, candidates code in Codespaces and submit a pull request, and Litebulb runs linter, integration and visual-regression tests, load tests, and complexity checks before compiling a report. Gary Lin positioned it as 'the depth of a service like Karat at the scale and price point of a tool like HackerRank', with a long-term ambition to become 'Webflow for interviews'. His replies named Gumroad, On Deck, Mashgin, Dover, SnapEDA, Evidence.dev, getatlas.io and Okteto as companies running Litebulb interviews. Much of the thread became a debate about whether automated assessment disrespects candidates, whether code metrics can judge people, and whether cheating is preventable; commenters also quoted pricing from $600/month up to enterprise tiers. Gary answered that Litebulb belongs mid-funnel with human screens before and after, and that a V2 scorecard would expose raw data instead of labels like 'strong junior'.
No ending yet — it is still running.
Background
Litebulb was a YC W22 startup automating coding interviews for remote teams. Founder Gary Lin, a software engineer who had worked at 11 companies and been through more than a hundred interview funnels, argued that LeetCode-style puzzles test the wrong skill and that building and running technical interviews pulls engineers away from real work. His bet was that the winning tool would automate the late-funnel stage that Karat served with humans and HackerRank skipped entirely.
The product made candidates do real work: access to an existing codebase with a database, server and client, Dockerized environments started with one make command, coding in browser-based Codespaces, submission as a GitHub pull request. Litebulb ran tests plus linter, visual-regression, load and complexity analysis and produced an employer report, with humans still verifying results during beta. Launched on Hacker News on 2022-03-07, the post drew 139 points and 190 comments, and the founder named Gumroad, On Deck, Mashgin, Dover, SnapEDA, Evidence.dev, getatlas.io and Okteto as users.
Reception was polarized. Hiring managers and candidates who hated the LeetCode grind praised the real-codebase approach; others called automated assessment impersonal and said numeric metrics create a false sense of objectivity. Gary countered that Litebulb was not a pre-screener, recommended human calls before and after, and was moving the scorecard from seniority labels to raw data. At launch it was still in beta, manually verifying results with a 24-hour turnaround, with data-science interviews, a candidate-facing platform and ATS integrations on the roadmap.
What has to be true
- Engineer time is technical hiring's biggest cost: employed engineers build, run and grade interviews, with Gary citing teams spending two hours a day or several onsite interviews a week on them.
- The market had an open lane: early-funnel screeners test brainteasers and Karat supplies humans at high cost, so a late-funnel, automated product at a tool's price point was a differentiated position.
- Real-codebase tasks resist the failure mode of puzzles: no optimal answer to copy, git-history diffing flags copied solutions, and prep means becoming a better developer, not a better LeetCode solver.
- The launch debate exposed the core risk: candidates and hiring managers distrusted automated judgment, and labels like 'strong junior' undercut objectivity, pushing Litebulb toward raw metrics.
What can be applied
Positioning against a hated status quo buys attention, but the product still has to prove its signal means something to both sides; Litebulb's launch fight was about trust in automation, not features.
Aftermath
As of 2022-03-07 Litebulb was still early: in beta for at least three more months, manually verifying every interview result with a 24-hour turnaround, but already serving paying customers including Gumroad, On Deck, Mashgin, Dover, SnapEDA, Evidence.dev, getatlas.io and Okteto. The roadmap included a V2 scorecard exposing raw data instead of seniority labels, employer self-service analysis, a candidate-facing practice platform later in 2022, data-science interviews, and ATS integrations like Greenhouse and Lever. Gary Lin told HN the biggest risk was execution, not market size.
Sources
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card