Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am an AI and machine learning engineer specializing in large language models, intelligent agents, retrieval-augmented generation, and production-grade ML services. I have built multi-agent systems, recommendation assistants, process mining solutions, and conversational applications using Python, LangGraph, Pydantic AI, LangChain, vector databases, structured outputs, guardrails, and observability tools. My experience also includes fine-tuning language models with LoRA, developing computer vision and NLP models, and deploying reliable services with FastAPI, Docker, MLflow, and CI/CD.…I am an AI and machine learning engineer specializing in large language models, intelligent agents, retrieval-augmented generation, and production-grade ML services. I have built multi-agent systems, recommendation assistants, process mining solutions, and conversational applications using Python, LangGraph, Pydantic AI, LangChain, vector databases, structured outputs, guardrails, and observability tools. My experience also includes fine-tuning language models with LoRA, developing computer vision and NLP models, and deploying reliable services with FastAPI, Docker, MLflow, and CI/CD.

Daria Ovchinnikova

AI Engineer, Data Scientist, AI Developer





I am an AI and machine learning engineer specializing in large language models, intelligent agents, retrieval-augmented generation, and production-grade ML services. I have built multi-agent systems, recommendation assistants, process mining solutions, and conversational applications using Python, LangGraph, Pydantic AI, LangChain, vector databases, structured outputs, guardrails, and observability tools. My experience also includes fine-tuning language models with LoRA, developing computer vision and NLP models, and deploying reliable services with FastAPI, Docker, MLflow, and CI/CD.…I am an AI and machine learning engineer specializing in large language models, intelligent agents, retrieval-augmented generation, and production-grade ML services. I have built multi-agent systems, recommendation assistants, process mining solutions, and conversational applications using Python, LangGraph, Pydantic AI, LangChain, vector databases, structured outputs, guardrails, and observability tools. My experience also includes fine-tuning language models with LoRA, developing computer vision and NLP models, and deploying reliable services with FastAPI, Docker, MLflow, and CI/CD.

Available to hire

I am an AI and machine learning engineer specializing in large language models, intelligent agents, retrieval-augmented generation, and production-grade ML services. I have built multi-agent systems, recommendation assistants, process mining solutions, and conversational applications using Python, LangGraph, Pydantic AI, LangChain, vector databases, structured outputs, guardrails, and observability tools. My experience also includes fine-tuning language models with LoRA, developing computer vision and NLP models, and deploying reliable services with FastAPI, Docker, MLflow, and CI/CD.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

LLM Engineer at REML

September 1, 2025 - Present

Built production-oriented LLM agents and RAG systems for banking, real estate, travel, process mining, and sports data use cases. Developed multi-platform agents with Pydantic AI, LangGraph, LangChain, OpenAI API, Milvus, Qdrant, MCP tools, hybrid search, structured outputs, guardrails, and tracing. Designed multi-agent process mining workflows with Streamlit and pm4py, including hypothesis generation, validation, reporting, and impact estimation. Optimized conversational costs through caching and model selection, implemented response-format controls, integrated corporate OAuth authentication, and established testing and CI/CD practices.

Machine Learning Engineer at Avito

May 1, 2024 - November 1, 2025

Developed and deployed ML inference services, including an event-driven LLM inference system using LoRA adapters, Kafka, ClickHouse, FastAPI, and MLflow. Fine-tuned language models with PEFT and DeepSpeed to extract key criteria from real estate agents’ calls. Organized manual and LLM-assisted data labeling, trained classical ML, NLP, and computer vision models, and developed a ViT model for room-type and image-render classification. Evaluated business impact through A/B testing and conversion analysis, identified sales-call inefficiencies, and conducted technical interviews.

Data Scientist at Sber

July 1, 2022 - May 1, 2024

Evaluated the GigaChat large language model, investigating looping behavior, sampling methods, hallucinations, summarization, and question-answering quality. Studied the relationship between Local Intrinsic Dimension and hallucinations and used clustering and topic modeling to identify domains for moderation data enrichment. Built regression, classification, NER, and computer vision models, compared results against baselines, analyzed stability and feature importance with SHAP and permutation importance, and developed a statistical testing library for synthetic data generation.

Computer Vision Engineer at Center for Cognitive Modeling, MIPT, Intelligent Transport Laboratory

January 1, 2022 - July 1, 2022

Developed a semantic segmentation training pipeline for robotic vision using small real-world datasets and larger synthetic datasets with PyTorch Lightning and WandB. Converted trained models to TensorFlow Lite for deployment on edge devices. Researched methods for reducing the number of segmented classes in a SegFormer transformer by inserting trainable class embeddings into the architecture.

Data Analyst Intern at T-Bank

October 1, 2021 - January 1, 2022

Built a regression model to predict optimal marketing campaign timing. Automated transaction-data processing for banking partners, including customer-flow and cohort analysis, and developed audience-selection rules for marketing campaign A/B tests.