About Diana

Timișoara, Romania
European Union

The human behind the systems.

Diana Draghici.

AI Software Engineer

I build AI agent infrastructure, LLM evaluation systems, and developer tools. My background combines production generative AI with Python, backend, and platform engineering.

Agent infrastructureLLM evaluationDeveloper productivity

Experience

What I work on

AI infrastructure.
Engineering in practice.

Current responsibilities · select an area to read more

AI Gateway & evaluation harnessesRuntime, benchmarking, output quality

Design and build remote and local harnesses for agentic software development. Evaluate model performance, benchmark stages, and the quality of generated code.

Engineering telemetryAdoption, Git activity, team-level insights

Instrument AI tool adoption, run stages, pull requests, and Git activity. Work with engineering managers to turn these signals into team-level insights.

Integrations & reusable workflowsAI tools connected to engineering work

Connect AI tools with GitHub, Bitbucket, Jira, and Confluence. Build orchestration pipelines, reusable skills, prompt libraries, and automation components.

Guardrails & internal toolsValidation, governance, supported infrastructure

Establish security review, data-handling guardrails, and output validation. Prototype dashboards, reporting automations, and code-intelligence tools; harden successful prototypes into supported infrastructure.

Skills

Follow the connections

Skills have a story.

Select one to see where it shows up in my work.

AI & infrastructure

Engineering & integrations

Pick a skill.
Find the experience behind it.

Also part of my background: production generative AI, backend/platform engineering, CI/CD, coding-assistant integration, prompt libraries, and reusable skills.