2025 · Cuee

Cuee — Self-Driving Hiring Platform

An all-in-one AI ATS that sources, screens, interviews, and ranks — your team just makes the calls.

TypeScriptBunReactPostgreSQLDrizzleRedisS3LLM orchestrationBullMQWebSockets

The role

CTO & Co-Founder — architecture & full build

Highlights

  • Architected a seven-service platform — ATS API, recruiter app, interview client, recording, async workers, a real-time AI interview agent, and a multi-channel bot — covering the whole hiring lifecycle.
  • Built "Scout", an AI that sources candidates, writes job descriptions in your house style, and screens every application against your rubric.
  • Built a stage-based automation engine: recruiters set rules in plain English ("advance anyone scoring 90+, email the rest"), and AI evaluates the condition and runs the actions — moving candidates, sending email or SMS, and firing steps before and after each AI interview.
  • Embedded an in-app AI assistant that operates the ATS through its own API as the logged-in user — extracted and open-sourced as agentside.
  • One codebase, two products: the full ATS (Cuee) and an interviewer-only edition (Airbase), switched by a single product-mode flag.
  • Cut recording and document-processing costs by ~90–99% with custom infrastructure in place of costly managed services.
  • Model-agnostic LLM orchestration with automatic provider fallback, plus GDPR data-privacy workflows built in.
Cuee

// inside_the_product

What Cuee does — and what I built

Scout is sourcing…
ALLinkedIn
GTGitHub
PFPortfolio

Scout AI

It sources, writes, and screens for you

Scout finds candidates, drafts job descriptions in your house style, and reads every application against your rubric — so a ranked shortlist is waiting before you open a single resume.

Applied
Screen
Interview
Offer

Applicant Tracking

Your entire pipeline, automated

Jobs, stages, candidates, and hand-offs in one place with the busywork automated — the full hiring workflow, not a point tool bolted onto an inbox.

scorecard8.4/10
Problem solving
Communication
System design

Evaluation

Every candidate on the same rubric

Structured scorecards across screening and interviews, so decisions rest on consistent, evidence-backed criteria instead of gut feel — and stay defensible.

AC4 contacts
NV2 contacts
BL6 contacts

CRM

Companies, contacts, and your team

A built-in CRM for client and candidate relationships, with roles and permissions for the whole team, so nothing about a relationship gets lost between hires.

whenAI interview completes
if score ≥ 90AI rule
move to Offer
send offer email

Workflow Automation

Automations you describe in plain English

Set rules on any stage the way you'd say them — "advance anyone scoring 90+ to Offer and email them; send everyone else a polite no." The platform turns that into an AI-evaluated condition and runs the actions on its own: move the candidate, send an email or SMS, and fire steps before and after each AI interview. The whole pipeline runs itself.

// architecture

The system at a high level

A high-level view — a TypeScript/Bun platform of focused services with AI agents embedded across the hiring lifecycle, model-agnostic and API-driven.

  1. 01

    Clients

    ReactWebSocketsService Worker

    One React app that ships as the full recruiter ATS or an interviewer-only edition, plus a real-time interview client with resilient client-side recording.

  2. 02

    Core ATS API

    BunPostgreSQLDrizzle

    The system of record for the whole hiring pipeline, with team permissions and GDPR data-privacy workflows built in.

  3. 03

    AI layer

    agentsideMulti-providerTool-use

    Scout (sourcing + JD generation), rubric-based scoring pipelines, and an in-app agent (agentside) that operates the ATS through its own API.

  4. 04

    Interviews & recording

    STT / TTS / VADTwilioS3

    A sub-second voice agent bridging web and phone, paired with a cost-effective custom recording service.

  5. 05

    Async & integrations

    BullMQRedisSlack / Teams

    Queue workers for the heavy AI jobs (parsing, scoring, transcripts) and a multi-channel bot for Slack and Teams.

What it is

Cuee is a self-driving hiring platform — a full applicant tracking system with AI woven through every stage. It sources candidates, writes the job descriptions, screens applications against your rubric, interviews people over video and phone, and hands recruiters a ranked shortlist. The team makes the decisions; the platform does the legwork.

I'm the sole technical founder. I designed the architecture and built the platform largely solo across TypeScript, Bun, and Python — a set of focused services with AI agents embedded throughout.

One platform, two products

The same application ships in two modes behind a product-mode flag: Cuee, the full ATS, and Airbase, an interviewer-only edition focused purely on the AI interview. One codebase and one deploy pipeline, two go-to-market products.

The hard parts

  • Scout, the sourcing agent. Finds candidates, writes JDs in your house style, and screens applications — an agent that produces real work, not a chatbot.
  • Automations in plain English. Recruiters write stage rules the way they'd say them — "move anyone scoring 90+ to Offer and email them" — and the platform evaluates the condition with AI and runs the actions (advance, email, SMS) before and after each interview.
  • An assistant that operates the ATS. It drives the product through its own API, as the logged-in user, inside their permissions. Open-sourced as agentside.
  • Interviews at production quality and cost. A real-time voice agent over web and phone, with custom infrastructure that cut recording cost by ~90%.
  • Decisions you can audit. Rubric-based scoring with structured, evidence-backed output across sourcing, screening, and interviews.
  • Never locked in. Model-agnostic orchestration with automatic fallback, and GDPR workflows built in.

Want something like this built?

~/work/cuee$