Talantiq: a white-label HRM platform for Kazakhstan and Uzbekistan

HR across the region lives in scattered spreadsheets, chats and off-the-shelf boxes that don't cover the full employee lifecycle and ignore local specifics — languages, HR record-keeping, personal data requirements.

What's needed is a single platform for the KZ/UZ market that can be deployed under each client's brand as a separate, isolated instance.

• Recruiting (ATS): job openings, candidates, hiring funnel, hh integration.
• Adaptation and onboarding: checklists, the employee's first day.
• Learning (LMS): courses, development tracks, skills inference.
• Performance appraisals: goals, reviews, assessments with a code runner.
• HR document flow: orders, employment contracts, approvals.
• Time and leave: the KZ/UZ working calendar, day calculations.
• Engagement: eNPS, surveys, points and gamification.
• Analytics: end-to-end metrics, dashboards, a notification center.

The assistant is built on RAG (vector search in Qdrant) and a pluggable LLM gateway: you can plug in a cloud model or a self-hosted one — to meet data residency requirements.

It doesn't just answer — it performs actions on data through tool-use, with guardrails on the output.

Each client gets an isolated instance with their own brand: theming, feature flags, tenant provisioning.

Above the fleet of instances sits the vendor's operator console (control-plane), which works with metadata and aggregates, without access to clients' personal data.

The backend is NestJS, Prisma and PostgreSQL, Redis for caching and queues, Qdrant for vectors; the web app on React, mobile on React Native, the career site on Next.js. Observability is OpenTelemetry, Prometheus, Grafana, Loki.

RBAC, audit logging and personal data retention are built into the architecture; instances are deployed on Kubernetes or on-premise, with data storage localized to meet KZ/UZ requirements.