Looking for a job

Gregory Harutyunyan Gevorgyan

@gregoryharutyunyange

AI / Machine Learning Engineer who migrated Qualígrafo’s AI assistant API to asynchronous HPC jobs and added flexible LLM inference.

Spain

Message

What I'm looking for

I'm available for internships until July 2027 through a university internship agreement.

At Qualígrafo S.L., I migrated the Mimesis Platform AI assistant API to an asynchronous job-based system, avoiding timeouts on long-running executions and validating the architecture for the SCAYLE supercomputing cluster. I also implemented JSON Schema and Pydantic validation in an orchestrated LLM pipeline, and developed a LangChain layer to switch between LLM providers.

For my MSc thesis at UNIR, I developed a machine-learning approach to predict quantum circuit noise before execution, with up to 81% error reduction. My projects also include comparing RAG and knowledge-graph approaches for university regulations, and building an explainable hate-speech classification pipeline.

Experience

Work history, roles, and key accomplishments

Backend AI Engineer / LLMOps

Qualígrafo S.L.

Mar 2026 - May 2026 (2 months)

Migrated the FastAPI backend of an AI assistant to an asynchronous job-based system, avoiding timeouts on long-running executions. Implemented JSON Schemas and real-time validation with Pydantic for 8 prompt stages in an orchestrated LLM pipeline, and developed an abstraction layer with LangChain to switch between 3 LLM providers.

fastAPI Pydantic Lang Chain Ollama GROQ Azure OpenAI JSON schema HPC Scylla Python

Education

Degrees, certifications, and relevant coursework

International University of La Rioja (UNIR)

Master of Science, Artificial Intelligence

2025 -

Pursuing an MSc in Artificial Intelligence with a Data Engineering specialization, including a master's thesis on pre-execution noise prediction for quantum circuits using machine learning and graph neural networks.