Looking for a job

Javkhlantugs Tsedensumiya

@javkhlantugstse

Machine Learning Engineer specializing in production LLM and speech AI systems with Computer Vission and Retrieval Augmented System.

Mongolia

Message

What I'm looking for

I’m looking to build production LLM and speech AI that’s reliable and fast—RAG pipelines, AI agents, and low-latency services with strong observability. I want to own end-to-end delivery from model workflows to Dockerized deployments.

I’m a Machine Learning Engineer focused on building production-ready LLM and speech AI solutions, especially scalable retrieval-augmented pipelines and AI agents. I design end-to-end services that take models from prototype to production with reliability and performance as core priorities.

In my work, I’ve built RAG systems with embedding methods and vector stores, developed multi-step AI workflows orchestrated through FastAPI, and automated content summarization and categorization using transformer-based language models and prompt engineering. I also deliver speech AI with Whisper STT and Piper/Kokoro-82M TTS, plus document understanding via OCR/vision models like Tesseract and LayoutLM—then package everything into Dockerized deployments and ML microservices optimized for low latency, observability, and runtime durability.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Engineer

Current

Independent AI Systems

Jan 2023 - Present (3 years 5 months)

Built production-oriented RAG and AI agent pipelines using local language models via Ollama, including PostgreSQL vector-store retrieval and end-to-end document/news summarization and categorization. Developed speech AI (Whisper STT and Piper/Kokoro-82M TTS) plus OCR/document understanding workflows, and deployed FastAPI services in Docker for reliable low-latency operation.

Retrieval Augmented Generation (RAG)Python fastAPI Docker Ollama MongoDB Whisper STT TTS (Piper Kokoro 82M)OCR (Tesseract LayoutLM)

Data Scientist

Petroleum Distribution Company

Jan 2022 - Jan 2024 (2 years)

Built end-to-end Python data analysis and reporting pipelines using Pandas and SQLAlchemy with PostgreSQL, delivering daily performance dashboards to support company decisions. Automated weekly reporting workflows with Streamlit and scheduled Python scripts to reduce manual effort and improve timeliness.

Python Pandas SQLAlchemy PostgreSQL Data Pipelines Dashboards Streamlit Report Automation Scheduled Jobs

Mobile Application Developer

Independent Development

Jan 2015 - Jan 2022 (7 years)

Developed Android (Java) and iOS (Swift) mobile applications using modular architecture and automated testing to improve stability and increase engagement. Built API-integrated mobile systems with Retrofit and Alamofire, improving data synchronization and overall user satisfaction.

Java Swift Android SDK Retrofit Alamofire API Integration Automated Testing Modular Architectures