AI Engineer (LLM, RAG, Agentic Workflows)

About the role

Volga Partners is hiring an AI Engineer in Islamabad, Pakistan, to design, develop and deploy advanced AI systems such as LLMs, RAG frameworks, agentic workflows and intelligent automation. The role reports to Product & Engineering Leadership, works in a hybrid arrangement (3 days in office, 2 days remote) and involves close collaboration with software engineers, product teams and leadership to deliver scalable AI solutions for global clients.

Key responsibilities

• Design, build and deploy AI systems, including LLMs, RAG, agentic workflows and ML/DL solutions, choosing models that balance accuracy, latency, cost and complexity.
• Train and fine-tune models with PyTorch or Keras, including data preparation and evaluation, and implement and benchmark research methods.
• Build APIs, inference pipelines and backend integrations.
• Own evaluation metrics, monitor and debug systems after launch, and ensure production readiness.
• Work with software engineers and product teams, document architecture and technical decisions, and read and adapt ML research papers.

Required profile

• 3+ years of hands-on AI/ML engineering experience, with at least one system deployed to production.
• Strong NLP fundamentals (tokenization, embeddings, classification, NER, information extraction).
• Ability to read ML research papers and turn them into working code.

Required skills

• Hands-on model training with PyTorch or Keras.
• Production experience with LLMs, RAG, embeddings/vector databases and prompt/evaluation design.
• Coding, Git, testing, API/system design and debugging.
• FastAPI or similar, Docker, and at least one cloud platform (AWS, GCP or Azure).
• Preferred: LLM fine-tuning (LoRA, RL-based), SQL, LangGraph/LangChain, Kubernetes, CI/CD, model-serving infrastructure, open-source contributions.

What we offer

• Monthly salary of PKR 150,000 to 400,000, depending on experience and expertise.
• Engagement through Volga Partners, with contract and payment handled via Deel.
• 90-day probation, then medical insurance, paid time off, holiday pay and career growth opportunities.

Hiring process: recruiter screening video call, asynchronous practical assessment, assessment review, and a final interview with the CTO.