Looking for a job

Ekaterina Prokhorova

@catprokhorova

Message

I build production RAG, search, and retrieval systems that improve accuracy and cut latency.

Serbia

Message

What I'm looking for

I'm looking to build production ML, RAG, search, and retrieval systems where architecture, measurable quality, and real-world infrastructure constraints matter.

I've rebuilt document search systems at InTech LLC and Technologies of Trust, cutting search latency from about 80 seconds to 2 seconds and improving retrieval Recall@K from 0.60 to 0.80. My work spans hybrid search, embedding fine-tuning, asynchronous RAG pipelines, and evaluation built around real business risk.

At Huntee, I lead ML architecture for a job and resume matching platform, defining production microservices for document parsing, bilateral matching, and skill extraction. I combine applied NLP with Python backend engineering to ship reliable services using FastAPI, OpenSearch, Weaviate, Docker, and CI/CD.

Experience

Work history, roles, and key accomplishments

Current

ML Lead

Current

Huntee

May 2026 - Present (3 months)

Lead ML architecture for an AI-powered job/resume matching platform, defining microservices and data contracts. Initiated GLiNER-based NER microservice and bilateral RAG matching architecture.

RAG LLM Microservices NER GLiNER Architecture Python fastAPI

ML Engineer, Backend Heavy

InTech LLC

Jul 2025 - Mar 2026 (8 months)

Restored document search system to production stability, cutting latency by 97.5%. Built hybrid search engine and image search using vector embeddings, and automated template classification.

OpenSearch Hybrid Search Embeddings Shingling TF IDF SVM SpaCy NER Python fastAPI

ML Engineer

Technologies of Trust

Jul 2023 - Jul 2025 (2 years)

Built RAG system for contract risk assessment, improving retrieval accuracy by 33% and cutting processing time by 90%. Developed news aggregation service with 1,000+ documents/day throughput.