We recruit engineers through work, not resumes.
We rank it against a published rubric.
What we're doing
We're building a category of business applications on the SUPERWISE platform with a small, sharp team. The work is consequential, the customers are real, and the timeline is now.
How we find the team
We don't recruit through resumes — we recruit through work. The exercise below is our standing invitation. When a role opens, top scorers whose experience matches move into our interview process. Other top scorers go on a list we come back to.
What we believe
The quality and velocity gap between engineers who've fully embraced agentic AI as their primary engineering substrate and engineers who haven't is the largest gap in software engineering right now. We're recruiting from the population that's already crossed it.
If you're driving Claude or an equivalent toolset every day — judging its output, directing its work, shipping production code at a velocity your employer can't keep up with — this page is for you. If you're using AI as autocomplete, it's not.
Currently Hiring
Staff Engineer, Product
We're recruiting a Staff Engineer to lead the build of our AI chatbot solution and the broader category of business applications it opens up. Reports directly to John Leschorn, CPO.
Experience
10+ years
Location
Indiana, Tennessee or surrounding states preferred
How to get hired
The Exercise
Build a Kanban application using Claude Code (preferred) or another agentic AI toolset.
The full brief, optional scaffolding, and our published rubric live in the starter repository:
github.com/superwise-ai/engineering-exerciseRead the README before you start.
What you submit
Public GitHub Repository
Containing your code. Public only — no zips, GitLab, Drive, or email.
AI Session Logs
Transcripts of how you built it, included in the repo.
Your CV
In the repo or linked from the README.
Current submission window
May 15 – June 15, 2026
We review all submissions after the window closes. Submissions received after the window closes are not reviewed.
When this window closes, we'll likely open another. If we have an active window posted above, the exercise is live. If not, you can still submit — we'll review when the next window opens.
What we're looking for
The Categories to be Scored
The full rubric is published in the starter repo. Weights are private.
The rubric is strict.
Leverage
How effectively you drive the AI to produce quality output.
Depth and complexity
Substance beyond surface CRUD.
Substrate-to-surface thinking
Reasoning across the full stack.
Security instinct
Catching vulnerabilities the prompt didn't ask about.
Creative nuance
Useful additions that demonstrate judgment.
What happens after
After the window closes, we review every submission and rank by rubric score.
Top 20 by score
- $25 gift card
- Feedback directly from John Leschorn, CPO, to thank you for the work.
CV Matches Role
- Move into the interview process for that role
- Each interview stage is a real gate, not advisory
Submissions
- We may ask permission to keep your submission on file for future consideration
A note on transparency: Submissions that don't pass the rubric receive no further communication. We're being explicit about that because the exercise attracts a wide range of candidates and we want to be honest about how we operate. The faster the wrong candidates self-select out, the better the experience for everyone.
Don't
Don't email us with questions about the rubric, the exercise, or the role before you submit. The submission is the conversation. After you've passed the rubric, we'll have plenty to talk about.
Don't submit early "draft" versions for feedback. We review once, after the window closes.
Don't ask for extensions to the submission window. There aren't any.
Ready to submit?
When your work is ready, submit your GitHub repo URL through our submission form. Public GitHub only.