The AI Recruiter That Never Forgets a Candidate
The call dropped. Mid-interview. A candidate had just answered the second question of a five-question phone screen. Then the connection died. On most call workflows, that's a restart: the next call starts at question one, the candidate repeats themselves, and the notes from call one are gone. That moment is why I built this. The code is here:…
The AI Recruiter That Never Forgets a Candidate is an innovative solution designed to improve the candidate screening process through Telnyx's voice layer. When a call is dropped mid-interview, this system ensures that the conversation continues seamlessly from where it left off. The core of this technology is a candidate screening call agent that uses AI to manage the interview process.
The agent initiates a call, runs the interview with five predetermined questions, scores the responses against a role rubric, and records the data in a durable actor that exists on Telnyx Edge Compute. This durable actor holds the candidate's profile, their answer ledger, scores, and a history of call events. The key advantage of this architecture is that it eliminates the need for a "resume-after-drop" feature, as the candidate's history is inherently preserved within the durable actor.
Every outbound call carries its own instruction sheet, including the list of questions already answered. When the call is resumed, the assistant reads this sheet and proceeds to the next unanswered question, ensuring a smooth continuation of the conversation. The AI Assistant is equipped with an Ultra voice, expressive mode, patient turn-taking, and premium voicemail detection.
If no candidate answers the call, the assistant leaves a short callback message and stops the call. The scoring mechanism is transparent and honest, with answers evaluated against a set of ideal signals and real content. If an answer is vague or empty, it is routed for human review. The recruiter receives a live transcript and a scorecard as the call progresses, along with recommendations on the next steps, such as advancing, reviewing, or seeking human intervention.
The system also supports inbound SMS, allowing candidates to confirm their interview status via text, and hooks into existing Applicant Tracking Systems (ATS) for seamless integration. By applying this architecture to various hiring processes, such as patient intake, lead qualification, insurance verification, and appointment confirmation, the system ensures that candidate histories are preserved across different interactions, creating a more personalized and efficient hiring experience.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
