Prometa.ai is a rapidly growing technology company developing Deep Learning-based recommendation systems and agentic AI solutions powered by Large Language Models (LLMs). We operate active projects across various industries and continue to strengthen our pioneering position in the field. We offer AI-powered products and consultancy services tailored for sectors such as finance, retail, healthcare, insurance, and telecommunications.

We are looking for a Research Engineer who will lead the design, prototyping, and validation of advanced AI techniques across our product and service development lifecycle. The ideal candidate will have a strong background in NLP and LLM technologies, deep hands-on expertise in experimentation and rapid prototyping, and the ability to drive innovation in close collaboration with multi-disciplinary teams.

Key Responsibilities:

• Prompt and Reasoning Experimentation: Design, refine, and evaluate complex prompts and reasoning chains for zero-shot and few-shot LLM applications.
• Fine-Tuning and Adaptation: Execute supervised fine-tuning or parameter-efficient adaptation (e.g., LoRA, QLoRA) of foundation models for task-specific domains.
• RAG Benchmarking: Evaluate document retrieval strategies, embedding model quality, chunking heuristics, and reranking mechanisms for knowledge grounding.
• Prototype Agentic Workflows: Design and validate agent flows that use memory, tool invocation, and multi-intent routing for conversational systems.
• Model Evaluation and Analysis: Use custom evaluation frameworks to assess performance across latency, quality, hallucination, safety, and cost trade-offs.
• Knowledge Sharing: Document best practices and findings, and collaborate with DS and AI Engineering teams to ensure successful downstream implementation.
• Cross-functional Collaboration: Translate abstract requirements from Data Scientists into research plans, and deliver technically validated components to AI Engineers.

Required Skill Sets:

• Deep Understanding of LLM Architectures and Attention Mechanisms
• Hands-on Experience with Hugging Face Transformers, PyTorch, PEFT/LoRA
• Prototyping Tools: LangGraph, LangChain, custom agentic logic
• Vector Search Evaluation and Embedding Models
• Prompt Engineering and Reasoning Chains
• Experiment Management (Weights & Biases, MLflow, custom logs)
• NLP Evaluation Techniques (BLEU, ROUGE, hallucination metrics)
• Python, Jupyter, and lightweight deployment environments
• Communication and Technical Documentation
• Applied Research Mindset(PhDc or PhD) with Iterative Development Skills

What We Offer:

• Opportunity to work on leading agentic AI platforms in Turkey
• A forward-thinking, innovation-driven, and growth-oriented organization
• Hands-on experience with scalable AI solutions across multiple industries
• Flexibility to work remotely
• Ongoing support for certifications and training for professional development