Cristian Cretu

- Designed the AI observability platform with Langfuse and OTeL to capture real-world usage data and quality statistics on the platform’s AI features, giving key stakeholders the data to drive product decisions instead of guessing.

- Owned end-to-end development of an AI reporting assistant with a Vue.js/TypeScript frontend and a Spring AI backend on Spring Boot, using a GPT-5.5 model, that lets

- Built a full-stack AI marketing platform for SMBs, owning the Python/FastAPI backend, tested with Pytest, with Pydantic schema validation on every API contract and agent output, and the SvelteKit/TailwindCSS component-based, responsive, i18n-supported frontend validated with Zod.

- Architected a multiservice Python/FastAPI backend, splitting fast CRUD APIs from longer, durable execution workflo

- Led a team of 3 building a Flutter mobile application for elderly users, funded by the UN Population Fund (UNFPA Moldova).

- Integrated OpenAI API together with BeautifulSoup4 scraping pipelines and Vector Stores to ground AI outputs.

- Built backend services with Node.js, Python, and Firebase Functions; used Firebase Realtime Database for low-latency data delivery.

- Managed communication be

- Built a TypeScript anomaly detection service (z-score evaluator, autocorrelation checks), tested with Jest, that catches 5 to 10 real anomalies a week in client referral programs, flagging malicious behavior before it can damage a client’s program.

- Built custom ETL pipelines in Scala on Apache Spark via Apache Zeppelin, converting raw client data into reports for their business logic

- Enhan

- Built a data visualization backend in Python and Node.js to help client teams track and interpret referral program metrics.

Pursuing a Bachelor of Engineering in Computer Software Engineering at the Technical University of Moldova, with an expected graduation in May 2026.