Automate Candidate Pre-Screening with a Custom Voice AI

Recruiting teams often start with SaaS interview platforms. These tools are easy to set up but rely on rigid scripts and basic keyword matching for scoring. If your rubric requires judging the context of an answer, a keyword-based system fails. A candidate who says "I'm not certified, but I have operated a forklift" might get incorrectly passed through a simple filter.

A regional logistics company trying to hire 50 warehouse staff from 400 applicants faces this problem. Using a SaaS tool at $5 per interview costs them $2,000 for the initial screen. Worse, the recruiter still has to listen to hours of audio because the keyword search for "forklift certified" misses candidates who say "I operated a stand-up reach truck." The tool creates more review work, defeating the purpose.

Building on raw telecom APIs like Twilio is the next logical step, but it just provides the phone call plumbing. You are still responsible for building the state machine for conversation flow, integrating a separate transcription service, and then piping that text to an LLM for scoring. This quickly becomes a full-scale software project that internal teams lack the time or specific expertise to build.

Syntora's engagement for voice AI pre-screening begins with a detailed discovery phase to understand your existing interview scripts and scoring rubrics. We would work with your team to define 5-7 critical qualifying questions, mapping them to a data model compatible with your Applicant Tracking System (ATS), such as Greenhouse or Lever. This ensures the generated data integrates directly into your workflow. Client input would involve providing current screening materials, access to ATS documentation and relevant API endpoints, and subject matter expertise on desired candidate profiles. The deliverables would include a fully deployed, custom voice AI pre-screening system, comprehensive technical documentation, and basic operational training. Typical build timelines for a system of this complexity, including discovery, development, and integration, range from 6 to 12 weeks, depending on the depth of ATS integration and the complexity of conditional logic.

Leveraging the Claude API for its sophisticated instruction-following and ability to return structured JSON, the core system would be a Python-engineered conversational state machine to manage the interview flow. We've applied similar Claude API-driven document processing pipelines in adjacent domains, such as financial document analysis, and the same robust pattern applies to candidate responses. When a candidate reaches the pre-screening stage in your ATS, an event or webhook would trigger an AWS Lambda function. This function would initiate the call, play the pre-defined questions, and stream the candidate's audio to a high-quality real-time transcription service. The resulting transcribed text would then be sent to a Claude API endpoint with a custom-engineered prompt designed to score the response against your specific rubric.

The Claude API would be configured to return a structured JSON object, including a score (e.g., 1-10) for each question, a confidence rating, and a concise summary of the interview. A FastAPI service would then be developed to securely write this parsed and scored data to custom fields within the candidate's ATS profile. This integration means your recruiters would access a complete, scored summary directly within their primary tools, streamlining their workflow.

The proposed architecture would utilize serverless infrastructure for scalability and cost-efficiency. Supabase would serve as the data store for call logs and to manage rubrics, allowing for non-technical adjustments to screening criteria. For operational visibility, all application logs would be written as structured JSON using structlog, enabling precise monitoring and alerting configuration, should specific thresholds be exceeded. This serverless design also contributes to a lean operational cost profile, making it a sustainable solution for high-volume recruitment.