Looking for a job

Александр Ярославский

@0005771

Message

I build production NLP and LLM systems, from RAG and fine-tuning to scalable serving.

Armenia

Message

What I'm looking for

I'm open to remote work and relocation, and available immediately. I'm looking to build and operate production NLP and LLM systems, spanning model evaluation, RAG, fine-tuning, scalable serving, and MLOps infrastructure.

I've built production NLP and LLM systems since 2022, including banking assistants, retrieval pipelines, model fine-tuning, and low-latency inference services.

At Dushanbe City Bank, I built an assistant using Llama 3 8B and Mistral 7B that automated approximately 82% of responses with 94% routing accuracy. I developed hybrid-search RAG across 1,500+ documents with PII masking and fine-tuned Mistral 7B on 12,000 dialogues, raising intent recognition accuracy from 83% to 91%.

At Alfa Tech, I moved ML and LLM logic from Airflow DAGs into reusable Kubernetes-packaged modules and owned LLM evaluation and MLOps/DevOps work. Previously at X5 Group, I delivered PyTorch and Hugging Face feedback-classification models that reduced manual processing by approximately 40%.

Experience

Work history, roles, and key accomplishments

Current

MLOps Engineer

Current

Alfa Tech

Aug 2025 - Present (1 year 1 month)

Moved ML/LLM logic from Airflow DAGs into reusable modules packaged for Kubernetes. Owned LLM evaluation modules and MLOps/DevOps, maintaining Airflow/Spark workflows and improving asynchronous batch reliability.