AI Engineer with 7+ years of experience across software engineering, data systems, analytics, automation, and production-grade intelligent applications.

My strongest fit is applied AI: turning ambiguous business problems into reliable systems that can be deployed, operated, measured, and improved. I work end to end across agents, RAG, MCP-based integrations, full-stack product engineering, data infrastructure, cloud delivery, observability, and enterprise automation.

I like building systems where AI is not just a chat box, but a useful layer over real workflows: documents, spreadsheets, APIs, internal knowledge bases, business rules, queues, dashboards, and human review loops.

• Building agentic systems with governed tool use, grounded retrieval, and auditable outputs.
• Designing RAG and document intelligence workflows for enterprise knowledge and operations.
• Creating AI-powered spreadsheet, reporting, and analysis tools with Python/pandas execution sandboxes.
• Integrating LLMs with platforms such as SharePoint, Power Automate, Salesforce, Confluence, and internal APIs.
• Shipping full-stack AI products with strong reliability, diagnostics, testing, and user-facing polish.

• LLM applications with tool calling, function execution, context engineering, and multi-model routing.
• Agent workflows using OpenAI, Anthropic, Gemini, LangChain, Google ADK, and MCP.
• RAG pipelines with embeddings, semantic search, pgvector, source grounding, and response validation.
• Conversational interfaces over enterprise documents, spreadsheets, APIs, and internal knowledge bases.
• Real-time and multimodal experiences with audio streaming, Gemini Live Audio API, Web Audio API, Three.js, and shaders.

• AI-native products with Next.js, React, TypeScript, Node.js, FastAPI, Tailwind CSS, Zustand, tRPC, and Electron.
• Backend services with asynchronous processing, queues, workers, REST APIs, file pipelines, and structured observability.
• Product interfaces for workflows that need review, traceability, dashboards, exports, and operational control.
• Production delivery with Docker, CI/CD, testing, logging, diagnostics, and failure handling.

• Data pipelines and analytics workflows with SQL, Python, PySpark, pandas, Power BI, DAX, PostHog, and Google Analytics.
• Cloud and platform work across Azure, AWS, Vercel, Cloudflare, Databricks, Data Factory, Azure AI Foundry, and storage services.
• Enterprise integrations with SharePoint, Power Automate, Salesforce, SAP BW, Confluence, WhatsApp Business APIs, and internal systems.
• Data modeling, indexing, performance tuning, and scalable PostgreSQL/Supabase architectures.

Local MCP server, Chromium extension, and Codex skill for AI-operated Microsoft Power Automate flows. It exposes a supervised tool surface for inspecting, editing, validating, running, reviewing, and reverting flows through a local bridge and browser-session context capture.

Tech: TypeScript, Node.js, MCP, Chromium Extension, Power Automate, Dataverse, HTTP APIs.

Governed AI workspace based on LobeHub, with internal agents connected to SharePoint, methodology knowledge bases, corporate spreadsheets, and internal APIs. The system turns fragmented corporate knowledge into a practical conversational interface for operations.

Tech: Next.js, React, TypeScript, LobeHub, agents, SharePoint, internal APIs, tool calling.

Asynchronous validation workflow for PPT/PDF status reports using generative AI, document parsing, business rules, queues, workers, Power Automate callbacks, and executive HTML/PDF response generation.

Tech: TypeScript, Node.js, OpenAI, RabbitMQ, PDF/PPT parsing, workers, REST APIs, testing, observability.

Agent for large and multi-file Excel/CSV analysis with workbook inspection, semantic profiling, relationship-key detection, reconciliations, aggregations, chart datasets, and CSV/XLSX exports. Execution runs in an isolated Python/pandas sandbox with limits, caching, and validation.

Tech: TypeScript, Python, pandas, OpenAI, sandbox execution, CSV/XLS/XLSX parsing, PostgreSQL, file storage.

Real-time voice application with bidirectional audio streaming, Gemini Live Audio API integration, and audio-reactive 3D visualization. Demonstrates multimodal interaction, procedural visuals, and agent experience design.

Tech: TypeScript, Three.js, Web Audio API, Gemini, Web Components, shaders.

RAG pipeline for analyzing and explaining financial transactions using corporate documentation, semantic retrieval, multi-model AI, caching, and observability for contextual responses.

Tech: Next.js, Supabase/pgvector, OpenAI, Anthropic, Gemini, Confluence API, TypeScript.

• I design AI systems around real workflows, not isolated demos.
• I care about grounding, traceability, evaluation, permissions, and operational reliability.
• I move comfortably across product, frontend, backend, data, cloud, and automation.
• I prefer small, well-instrumented systems that can evolve safely over time.
• I like turning messy requirements into tools that people can actually use every day.

Senior AI Engineer, Applied AI Engineer, AI Product Engineer, AI Deployment Engineer, AI Platform Solutions Engineer, Senior Full Stack Engineer with AI focus, or Founding AI Engineer.

• LinkedIn: linkedin.com/in/mikaelluca
• GitHub: github.com/kaael1
• Email: mikaelluca1@gmail.com