
HRLLM / OCR
Task. The HR department processed resumes in dozens of formats (.rtf, .doc/.docx, .pdf, scans). Bringing each resume to the corporate template took up to an hour of manual work.
Solution
We implemented automation on the AIHub platform: an AI model with OCR extracts data from any file format, structures it and generates a standardized resume in the corporate template. HR only reviews the result.
Results
Technologies: AIHub, LLM, OCR (Tesseract/PaddleOCR), FastAPI.
Project stages
- Process mapping (1 week) — recruitment funnel and screening-criteria analysis.
- Model setup (2–3 weeks) — CV parsing, ranking and multilingual summarisation.
- Integration (1–2 weeks) — ATS/HRM connection and recruiter workflows.
- Calibration — quality review with HR, bias and accuracy controls.
Where this applies
Resume automation helps HR teams at banks in Astana, IT companies in Bishkek, and retail chains in Tashkent hiring at scale. See AI systems & chatbots.