Islam Asylbaev
Senior Backend/ML Engineer
New to Contra
Islam is ready for their next project!
AI-MAX — Enterprise RAG & LLM Assistant Platform
Enterprise AI platform designed to automate knowledge access and Service Desk workflows using Retrieval-Augmented Generation.
I designed the RAG architecture with Qdrant vector search, embeddings, reranking, automated ticket classification, and AI-generated response suggestions. The platform integrates multiple local and OpenAI-compatible LLM backends through a secure API gateway and supports multilingual queries in Kazakh, Russian, and English.
My work included RAG pipeline design, LLM integration, retrieval quality, response validation, security controls, usage limits, and production-oriented API architecture.
1
17
Field Workforce Tracking & Route Management Platform.
Enterprise backend platform for tracking field employees, routes, GPS positions, and work activity.
Built with Java, Spring Boot, PostgreSQL/PostGIS, and a microservice architecture. The system processes GPS tracks, detects visits to control points, calculates time spent on the road and at stations, and generates electronic route-sheet PDFs with daily and night-time metrics.
I worked on backend services, geospatial logic, inter-service integrations, route calculations, reporting, and production-ready API development.
0
11
Face ID & Helmet Detection System
Computer vision system for employee access control at industrial and construction sites.
The solution combines Face ID, liveness detection, and PPE/helmet recognition to verify that the person entering the facility is a real registered employee and is wearing the required safety helmet.
Built with Python, PyTorch, YOLO, OpenCV, and face recognition technologies. Designed for real-time camera processing and automated safety checks at site entrances.
0
14
Real-time intelligence for your entire workforce.
SWMS connects smart helmets, mobile and desktop applications, and a cloud command center into one AI-powered platform — so you always know where your people are, that they are safe, and how work is progressing.
0
25