Recruiting & Sourcing
helio / recruiting-sourcingGive it a job description and your rubric. It sources candidates, screens each one against the requirements, and hands back a ranked shortlist with the scoring rationale for every name on it.
Source and screen for this role: Senior Backend Engineer, 5+ years in Go, has shipped a payments or billing system, based in or willing to relocate to Berlin. Score each candidate against those requirements and return a ranked shortlist with the reason for each score. Ask me what you need before you start.
Swap in your own JD and requirements. The last sentence is what stops it screening on assumptions instead of your rubric.
What is Recruiting & Sourcing?
A teammate that turns a job description into a scored candidate shortlist: it sources names, screens each against the requirements, ranks them, and writes the reason behind every score. It ranks and explains; the hiring decision and the outreach stay yours, and it never auto-rejects on protected attributes.
What it does
Any chatbot screens a resume once. Helio runs sourcing as a standing job — from your context, kept current, delivered to you.
| Owns the job, not just one screen | It takes the JD, sources candidates, and returns a scored shortlist with a written rationale for each name. |
| Runs again without starting over | Saved once, it re-runs as new candidates appear and against each new req, not on a fresh prompt. |
| Works from your context | It reads the JD, your scoring rubric, and past hires in Notion, so it screens for the bar your team actually sets. |
| Keeps itself current | When the role or the market shifts, it re-scores the pool against the new requirements instead of waiting to be re-taught. |
| Lands where you work | The ranked shortlist and its rationale arrive in Slack, Lark, or email, ready for the hiring owner to act on. |
Three things it gets right
Every score carries a reason
"5 years Go, shipped a billing system, no payments experience." A written rationale per candidate you can audit, not a bare number.
Scored against your rubric
It screens on the requirements in your JD and your past hires, not a generic template of what a good candidate looks like.
It flags where it's unsure
A candidate it can't fully verify is marked uncertain, not padded into the list to hit a count.
What Recruiting & Sourcing connects to.
Questions
- Does it make the hiring decision?
- No. It ranks candidates and explains each score, and the decision stays with you. It won't advance, reject, or contact anyone on its own — the shortlist is a starting point for your call, not a verdict.
- How does it avoid bias?
- It scores against the requirements in your JD and never auto-rejects on protected attributes like age, gender, or ethnicity. Because every score carries a written rationale, you can read exactly what it weighed and correct the rubric where it's wrong.
- Can it source from a specific site?
- Yes. Name the site and it sources from there through the browser, reading public profiles the way a person would. Point it at more than one and it screens the combined pool against the same rubric.
- What does the shortlist actually contain?
- A ranked list of candidates, each with a score, the requirements they meet and miss, and the reason behind the score. Where it couldn't verify something, that candidate is flagged uncertain rather than scored on a guess.
From a JD to a scored shortlist.
Run the prompt above. It asks what it needs, sources the pool, and hands back a ranked shortlist with the reason behind every score.
Ask about this automation