NB//SYS
AVAILABLE FOR AI-DRIVEN PROJECTS
NIKITA
BABKIN
Senior ML / LLM Engineer
Building production AI systems, RAG platforms, LLM agents and intelligent products.
INITIALIZING NEURAL MATRIX...
NEURAL LOAD0.0%
01 / ABOUT
Engineering intelligence
I design AI products that survive contact with production — from data and evaluation to reliable inference and observability.
LLMRAGAI AgentsLangGraphKubernetesFastAPI
DATA
RETRIEVE
REASON
SHIP
02 / CAPABILITIES
Systems I build with
ML & NLP
PyTorchHugging FaceBERTspaCyCatBoostOpenCVBackend & Data
PythonFastAPIPostgreSQLTrinoKafkaAirflowSearch & Storage
QdrantOpenSearchS3RedisFAISSBM25Cloud & MLOps
KubernetesDockerYandex CloudMLflowLangfusePrometheus / Grafana03 / PROJECTS
Selected intelligence
FEATURED SYSTEM
AI Assistant
Multi-agent workspace that plans, reasons and executes complex business workflows.
LangGraphFastAPIOpenAI
FEATURED SYSTEM
RAG Platform
Enterprise knowledge retrieval with hybrid search, reranking and evaluation loops.
QdrantPythonKubernetes
FEATURED SYSTEM
Trend Detection
Real-time signals engine that turns noisy data streams into actionable trends.
KafkaNLPPyTorch
FEATURED SYSTEM
Knowledge Assistant
Secure conversational access to distributed corporate knowledge.
RAGPostgreSQLLLM
04 / EXPERIENCE
Where I made impact
Mars
Senior AI / ML / LLM Engineer
- Build, deploy and operate production LLM systems in Yandex Cloud and Kubernetes using Docker and vLLM.
- Created a creative AI assistant for about 5,000 users using multi-agent workflows and RAG.
- Delivered an enterprise knowledge assistant integrated with Qdrant, S3 and OpenSearch.
- Designed and productionized a multi-agent trend analytics platform with Streamlit, SQL, Trino and agentic loops.
- Design multi-agent and RAG service architectures, including C4 L1/L2 diagrams.
LangGraphLangChainQdrantFastAPIKubernetesvLLMLangfuse
MTS Digital
Machine Learning Engineer / AI Engineer
- Created an automatic parser and conversational interface powered by LLMs.
- Deployed models and AI services to MTS Cloud with Docker and vLLM.
- Built an end-to-end product matching system with hybrid and vector search, orchestration and monitoring.
- Developed a Kafka-based microservice for named-entity extraction from HTML documents.
FastAPIOpenSearchQdrantBGE-M3CatBoostAirflowMLflowKafkaspaCy
Lad IT
Machine Learning Engineer / Data Engineer
- Built recommendation, semantic search, real-time price monitoring and document processing systems.
- Developed NLP models for IT vacancy seniority classification using BERT, NLTK, scikit-learn and CatBoost.
- Created document extraction pipelines, database requirements and automated ER diagrams.
- Developed computer-vision prototypes for pose classification and grain image classification.
- Built and maintained data collection and ML orchestration pipelines with Scrapy and Airflow.
PythonBERTNLPFAISSBM25OpenCVMediaPipeTensorFlowAirflow
05 / STACK
Technology utilization
PythonFastAPIDockerKubernetesYandex CloudvLLMLangChainLangGraphOpenAIQdrantOpenSearchPostgreSQLS3TrinoKafkaAirflowMLflowLangfusePrometheusGrafanaPyTorchHugging Face
5+Years Experience
20+AI Projects
5000+Users
100+LLM Experiments
06 / CONTACT
Let’s build future together
Have a hard AI problem or an ambitious product? I’m always open to a thoughtful conversation.
Let's talk