Mark Sokolov

Design and deliver end-to-end AI automation pipelines for clients across multiple engagements. Build private, on-premise LLM integrations using Ollama and automate multi-step workflows with n8n.

Delivered a production-grade Java Spring Boot backend using strict TDD, with all logic validated using Mockito. Built a full CI/CD pipeline with GitHub Actions and managed database schema evolution through Liquibase migrations.

Scoped and delivered a complete high-availability web infrastructure solution using Nginx and HAProxy, eliminating single points of failure. Benchmarked the system under load with JMeter and deployed Java Spring Boot applications on AWS EC2 with CloudWatch monitoring.

Scoped requirements and delivered a production-ready data pipeline for a fintech client, automating ingestion and processing of 100,000+ banking transactions. Integrated a locally hosted LLM via Ollama to keep sensitive data on-premise.

Delivered a complete, fully playable game product from client brief to launch using Claude AI end-to-end. Built the entire product through AI orchestration and prompt engineering, managing the full delivery lifecycle independently.

Standardized local development environments and rolled out Okteto cloud development environments. Partnered with engineers to troubleshoot environment and deployment issues and ran end-to-end testing of database operations.

Pursued a Bachelor's degree in Computer Science, completing a thesis on high-availability web systems using Nginx, HAProxy, AWS, and performance testing.