AI Researcher / Data Scientist

LLMs hold great promise but today making them reliable and improving their accuracy requires a lot of manual effort. At Log10, we are focussed on solving this challenge. We are looking for an AI Researcher / Data Scientist to help build self-improving LLM applications. You will work directly with the founders and the AI & engineering teams to make LLMs usable in many different professional scenarios.

Responsibilities

• Develop and implement novel AI algorithms to automate improvements of LLM applications
• Train fine-tuned and reward models based on large datasets to reflect human preferences of domain experts
• Invent synthetic data generation algorithms to train custom evaluation models and augment human review
• Collaborate with cross-functional teams to integrate AI solutions into our LLMOps product offerings
• Stay updated with the latest trends and developments in LLMs, especially evaluations, prompt engineering, fine-tuning and open source models (DBRX, Mistral, Mixtral, Llama-3 etc.)
• Document and present research findings and methodologies to both internal and external stakeholders

Requirements

• Master’s or PhD in Computer Science, Machine Learning, Artificial Intelligence, or a related field
• Proven experience with machine learning algorithms and statistical models. Solid understanding of NLP and experience working with LLMs.
• Strong programming skills in Python and experience with AI/ML frameworks like TensorFlow, PyTorch, etc.
• Excellent problem-solving and analytical skills
• Ability to work independently in an organized manner. Strong communication and collaboration skills
• Comfortable working with version control (git)
• Significant post-academic experience (will consider candidates at multiple levels)

Nice To Have

• Experience in a startup environment
• Experience with evaluating LLMs, familiarity with research in LLM evaluation, e.g. Prometheus
• Experience in logging, debugging and fine-tuning large language models
• Experience developing end-to-end machine learning systems
• Publications in relevant AI/ML conferences and journals
• Experience with cloud computing platforms like AWS, GCP, or Azure
• Knowledge of containerization and orchestration technologies (e.g., Docker, Kubernetes)