Inspiration
Every hiring manager we spoke to said the same thing — they spend more time reading resumes than actually hiring. We wanted to fix the most painful, repetitive part of enterprise recruiting with AI that actually works.
What it does
TalentScreen AI is a 3-node Airia active agent pipeline. Drop in a resume and job description — Agent 01 parses the candidate profile, Agent 02 scores fit from 1-10 with reasoning, Agent 03 drafts a personalized interview invite or rejection email. The hiring manager approves at the human-in-the-loop gate before anything sends. Under 10 seconds, start to finish.
How we built it
Built entirely on Airia's Agent Builder with 3 connected LLM nodes running in sequence. The frontend demo supports PDF upload, live scoring, and email drafting — all calling the Airia pipeline via API.
Challenges we ran into
Getting the scoring agent to return consistent structured output across edge cases took multiple prompt iterations. Tuning the human-in-the-loop gate to feel natural rather than friction was also key.
Accomplishments that we're proud of
A fully working end-to-end pipeline on Airia that any recruiter can use today with zero technical knowledge. The HITL gate keeps humans in control while AI does the heavy lifting.
What we learned
Airia's Agent Builder makes multi-step AI orchestration accessible without backend infrastructure. Prompt design between nodes matters more than model choice.
What's next for TalentScreen AI
Slack and Outlook integration so candidates get screened the moment their email lands. Bulk processing for high-volume hiring teams. ATS integrations with Greenhouse and Workday. And also a AI virtual meet to check their dress code during interview
Built With
- airia-agent-builder
- airia-llm-nodes
- airia-pipeline-api
- css
- html
- javascript
Log in or sign up for Devpost to join the conversation.