Mykola M.

Full-Stack AI Lead with 15+ years of experience in AI/ML/Data Engineering. Specialist in productionizing Document AI, Agentic LLM Workflows, and Scalable Big Data systems. Master’s in Applied Mathematics. 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈 & 𝐋𝐋𝐌 - Architecting production-grade agents and multi-intent workflows using LangGraph and LangChain. - Implementing NeMo Guardrails, DeepEval, and Langfuse to ensure security and near-zero hallucination rates. - Developing high-performance FastAPI backends for real-time operational data integration. 𝐃𝐨𝐜𝐮𝐦𝐞𝐧𝐭 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 & 𝐎𝐂𝐑 - Specialist in extracting structure from "dirty" or non-digital sources using Visual NLP and OCR. - Lead Developer of Spark OCR. - Author and main contributor of open-source projects: Spark PDF and ScaleDP. - Expertise in document layout analysis, data extraction, and semantic search. 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧 𝐌𝐋𝐎𝐩𝐬 & 𝐁𝐢𝐠 𝐃𝐚𝐭𝐚 - Architecting scalable infrastructures using Vector DBs (Milvus, Pinecone, Qdrant). - Optimizing Apache Spark (PySpark, Streaming) and Databricks pipelines; reduced processing time from days to hours. - Implementing robust CI/CD via Databricks Asset Bundles, Jenkins, and GitHub Actions. 𝐏𝐫𝐢𝐯𝐚𝐜𝐲 & 𝐂𝐨𝐦𝐩𝐥𝐢𝐚𝐧𝐜𝐞 - Automated de-identification and redaction for GDPR/HIPAA. - Specialist in masking sensitive information in DICOM, PDF, and image formats. 𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 AI & LLM: LangGraph, LangChain, RAG, Llama 3, Gemini, GPT-4, Vertex AI, Hugging Face, GliNER. Backend: Python, Scala, FastAPI, Node.js, PostgreSQL, MongoDB. Vector DBs: Pinecone, Milvus, Qdrant. Infrastructure: AWS, GCP, Azure, Docker, Kubernetes, Jenkins, AirFlow. Big Data: Apache Spark (PySpark, Streaming, MLlib), Kafka, Kinesis.