Looking for a job

Soha Pervez

@sohapervez

Message

I build production AI systems, RAG pipelines, and LLM safety tooling that make capable AI safer.

Finland

Message

What I'm looking for

I'm looking for a fast-moving team where I can ship real AI products, build capable and safe systems, and help get AI right.

I've shipped production AI systems at GPT-Lab, Tampere University, including a public RAG platform for AI safety research and an automated LLM red-teaming pipeline.

I built the EVIL-AI RAG System end to end, from PDF ingestion, hybrid retrieval, and reranking to streaming LLM responses on OpenShift. The platform made AI safety research publicly accessible without manual curation overhead.

My work also includes teaching master's students about LLM alignment, safety, and jailbreaking, and presenting AI safety findings to wider audiences. At HMD Global and Finnos, I turned large datasets into product, risk, forecasting, and production improvements using machine learning, statistical testing, NLP, and visualization.

I'm looking to join a fast-moving team that ships real products and cares about getting AI right.

Experience

Work history, roles, and key accomplishments

AI Engineer (Researcher)

GPT-Lab, Tampere University

Sep 2025 - Mar 2026 (6 months)

Designed and deployed a production RAG pipeline on OpenShift, enabling public access to AI safety research. Built an automated LLM evaluation pipeline and delivered a master's-level lecture on LLM alignment and safety.

RAG LLM Evaluation OpenShift N8n Python Docker Kubernetes AI Safety