← Home
Experience
Where I've been.
Four years across AI engineering, data science, and running marketplace shops — moving steadily toward production AI systems that actually ship.
-
AI Engineer · SinoAI
- Developed and deployed a production AI medical consultation chatbot across Telegram, web, and mobile with multilingual support.
- Built a custom medical RAG pipeline on ChromaDB + LangChain with a 4-step synthesis flow (Describe → Classify → Search → Generate).
- Shipped real-time voice consultation: VAD, STT, LLM streaming, and TTS over WebSocket.
- Engineered intent classification, emergency symptom detection, specialty routing, and multi-modal analysis (X-rays, PDF records).
- Designed 15+ Tortoise ORM / PostgreSQL models; set up Langfuse observability; containerized with Docker on GCP and AWS.
- Led AI-tooling adoption across the engineering team.
-
Data Science Intern · Aisha AI
- Collected and organized audio-text datasets for Uzbek TTS and STT model development.
- Cleaned and preprocessed data to improve quality and consistency across modalities.
-
Project Manager · Data Analyst · CommerceHub
- Ran six online marketplace shops with a primary focus on Uzum Market.
- Kanzec: grew monthly revenue from 80M to 160M UZS in 2 months, stabilized at 150–160M.
- Bikon.uz: grew monthly revenue from 60M to 230M UZS in 3 months.
- Built data-driven strategies for market analysis, product recommendations, and promotional campaigns.
-
Data Science Intern · RealSoft AI
- Applied Python, NumPy, and Pandas to clean and analyze real-world datasets.
- Built classification and regression models with scikit-learn, including a house-price regressor.