Actively Searching
Research Scientist
John Doe
Applying multimodal generative models and rigorous evaluation to build reliable AI for biology.
I am a research scientist focused on multimodal generative AI, model evaluation, and human-AI interaction. With experience at OpenAI and DeepMind and a Ph.D. track at Columbia, I build scalable evaluation and training pipelines to improve model reliability and translate LLM advances into deployable solutions for biology.
Currently
Leading multimodal LLM research and evaluation to deploy reliable biology AI.
Résumé
DownloadSkills
Languages
Python
C++
Java
SQL
JavaScript
Frameworks
PyTorch
TensorFlow
Hugging Face
LangChain
Cloud & Tools
AWS
Docker
Kubernetes
Linux
Git
Research Areas
LLMs
Multimodal Learning
Generative AI
NLP
Reinforcement Learning
Computer Vision
Retrieval-Augmented Generation (RAG)
Long-context Reasoning
Human-AI Interaction
AI Alignment
Infrastructure & Systems
Distributed Training
Scalable Evaluation Pipelines
Kubernetes (distributed training)
Methods
Model Evaluation & Reliability
Hallucination Reduction
Experimentation & Large-scale Model Experiments
Production & DevOps
Model Deployment
Professional
Mentoring
Cross-functional Collaboration
Presentation & Stakeholder Communication
Research Publication
Looking ahead
I apply multimodal generative models and rigorous evaluation methods to real-world problems, currently focusing on biological AI. I aim to translate LLM advances into reliable, deployable solutions for biology.
Connect
Featured
Multimodal LLM Evaluation Framework
Reduced hallucination rates by 22% and improved overall model reliability by building an evaluation framework tailored to multimodal large language models. Integrated the framework into production workflows via collaboration with product, safety, and infrastructure teams to monitor and benchmark model behavior.
Python · PyTorch · Hugging Face · LangChain · AWS · Docker · Kubernetes
Distributed Transformer Training Pipeline
Enabled training of large-scale transformer models by designing and deploying distributed training pipelines that scaled across cluster resources. Used the pipeline to support multimodal document-understanding research and to accelerate experimentation for graduate research projects.
PyTorch · Kubernetes · Python
Experience
Research Scientist
July 2028 – Present
OpenAI
Improved model reliability and reduced hallucination rates by 22% by designing and implementing robust evaluation frameworks. Deployed those evaluation systems into production through cross-functional collaboration with product, safety, and infrastructure teams and mentored junior researchers and interns on large-scale model experimentation and AI alignment practices.
Graduate Research Assistant
Aug 2023 – May 2028
Columbia AI Research Lab (Columbia University)
Published papers at top-tier conferences including NeurIPS and ACL, advancing multimodal document understanding and knowledge retrieval. Implemented distributed training pipelines for large-scale transformer models using PyTorch and Kubernetes and collaborated with faculty and graduate researchers on human-AI interaction studies.
AI Research Intern
Summer 2026 – Summer 2026
Google DeepMind
Enabled large-scale evaluation of generative AI systems by building pipelines that processed millions of samples daily. Ran experiments on retrieval-augmented generation and long-context reasoning and communicated findings to senior research leadership and engineering stakeholders.
Education
Ph.D. in Computer Science
– May 2028
Columbia University
Research Focus: Generative AI, Multimodal Learning, Human-AI Interaction
B.S. in Electrical Engineering and Computer Science
– May 2022
Massachusetts Institute of Technology (MIT)
GPA: 4.8/5.0