Open to opportunities

Igor Polubarev

@igorpolubarev

Message

AI solution architect and engineer building secure agentic systems, multilingual RAG, and Text-to-SQL for enterprises.

Israel

Message

What I'm looking for

I’m looking to lead and build secure, enterprise-ready agentic AI platforms—turning stakeholder goals into POC-to-production delivery with strong governance, auditable integrations, and measurable quality improvements.

I’m an AI engineer and solution architect with 7 years across engineering, data, and AI, including 4+ years designing and delivering enterprise AI systems. I build agentic workflows, multilingual RAG, Text-to-SQL, and Document AI for regulated environments, taking solutions from POC to production while keeping security and reliability central.

At IBM, I lead the architecture of a hybrid-cloud agentic platform connecting cloud-based agents to on-prem Jira, Confluence, GitHub, and internal APIs—preserving security boundaries and data sovereignty. I’ve designed on-prem multi-agent orchestration, improved multilingual RAG relevance by 70%, reduced NL2SQL query response time by 60%, and cut OCR+LLM document review time by 50%, guiding discovery workshops and translating business constraints into deliverable AI roadmaps.

Experience

Work history, roles, and key accomplishments

Current

AI Solution Architect

Current

IBM

Jan 2026 - Present (6 months)

Led the architecture of a hybrid-cloud agentic platform connecting cloud agents to on-premises enterprise systems while preserving security boundaries and data sovereignty. Defined reusable MCP and tool-integration patterns for controlled, auditable access to enterprise APIs and governed data sources.

Hybrid Cloud Architecture AI Agents Agentic Systems Workflow Orchestration Enterprise Integration Evaluation With Stakeholders

AI Engineer / Solution Architect

IBM

Sep 2023 - Dec 2025 (2 years 3 months)

Designed and deployed an on-premises multi-agent system orchestrating multi-step workflows across internal APIs and enterprise tools. Built multilingual RAG, an agentic NL2SQL system, OCR+LLM pipelines for regulated document understanding, and an LLM-assisted Angular prototyping workflow.

AI Agents Workflow Agents Multi Agent Systems Tool Function Calling Embeddings Semantic Search Milvus NL2SQL LLM Evaluation LLM As Judge Human Review OpenSearch PostgreSQL Angular MLFlow CI CD

Data Scientist

MNO Go Games

Aug 2022 - Aug 2023 (1 year)

Built machine learning models for question correctness prediction and adaptive question recommendation. Conducted product and traffic analytics using statistical testing and created dashboards for decision-making.

CatBoost Question Similarity Modeling Model Evaluation SVD++Neural Collaborative Filtering Time Series Analysis Cohort Analysis Clustering A B Testing Tableau Grafana

Engineer / Data Analyst

Boeing

Jun 2019 - Apr 2022 (2 years 10 months)

Analyzed engineering performance data to identify recurring major error patterns and support process-improvement decisions. Applied regression and classical ML to estimate engineering labor time and maintained documentation and stakeholder communication across multinational teams.

Data Analysis regression Machine Learning Performance Engineering Stakeholder Communication Documentation