Screening resumes against your own criteria with AI
A single hiring round can bring in dozens of resumes, and reading every file to find the few right candidates eats a whole afternoon. The Recruiting bot in Piskes Agent reads every resume in a folder you point it to, summarizes and ranks candidates against the criteria you set, then drafts interview invitations for the ones who fit.
What the bot does
- Role: HR & hiring. Screens resumes by your criteria, summarizes each candidate, drafts interview invitations for your approval.
- Skills: file reading, web search (to look up public information such as a portfolio, when useful).
- Connector: Gmail (to draft interview invitations).
- No schedule — used per hiring round.
Set up in 10 minutes
- Install Piskes Agent and configure a model provider.
- Marketplace → Bots → Recruiting → Add.
- Gather resumes (PDF/DOCX) for one hiring round into a folder, e.g.
D:\Hiring\accountant-2026-09. - Message it: “Resumes are in this folder. Criteria: at least 2 years of accounting experience, advanced Excel, a certification is a plus.” It reads them and returns a ranked table with reasons.
- For the fits, ask “draft interview emails for the top 3 candidates” — it produces drafts for you to review and send.
What happens to your data
- Resumes stay in the folder on your machine; only what’s needed to summarize and rank is sent to the model you chose.
- Sending an interview invitation affects someone else — at the Ask first level, the bot only drafts and stops to ask before sending.
- Candidate information is sensitive personal data — consider a local model if your hiring process requires resumes to never leave the machine.
Tips
- Spell out disqualifying criteria (“skip resumes without contact info”) for tighter filtering.
- For a high-volume round, split candidates by week rather than dumping everything into one folder, so the bot works faster.
- Once interviews are scheduled, use the Admin assistant bot to help manage the follow-up calendar.
Cost: on the free plan with your own key, screening and ranking around 20 resumes usually costs a few tens of thousands of tokens depending on resume length.