About Experience Open Source Research Contributions Press Contact
Senior Research Engineer

Teja Gollapudi

Post-Training & Agentic Large Language Models at Meta

Building the next generation of LLMs through reinforcement learning and post-training, inducing agentic reasoning, tool-use, and factuality in foundational models deployed to hundreds of millions of people worldwide.

Advancing AI at Scale

I am a Senior Research Engineer at Meta, where I explore post-training research on foundational language models. My work focuses on inducing and enhancing agentic capabilities in LLMs through reinforcement learning, advancing how AI systems reason, plan, and act across multi-turn, multi-tool, and multi-modal environments.

Beyond research, I have shipped production AI features for Wearables at scale. I contributed to the launch of Ray-Ban Meta smart glasses, Ray-Ban Meta Displays, and Oakley sport glasses, delivering real-time sports experiences across major sporting events and bringing Meta's Fitness AI features to millions of users worldwide.

Prior to Meta, at VMware I co-led the development of commercially viable open-source instruction-following LLMs, now downloaded tens of thousands of times monthly on Hugging Face. I also architected the LLM API infrastructure that serves over 2,000 daily requests across 40+ internal teams.

I hold an M.S. in Computer Science from the University of Illinois at Chicago. My research has been published at ICML, EMNLP, and IEEE, and featured in IEEE Spectrum, The Register, and Business Insider. I am committed to pushing the frontier of AI.

Computer Vision LLMs VLMs NLP Reinforcement Learning AI
Teja Gollapudi
Senior Research Engineer, Meta

Experience & Education

Senior Research Engineer
Meta
Industry
Machine Learning Researcher / Engineer
VMware
Industry
M.S. Computer Science
University of Illinois Chicago
Education

Models & Datasets

Co-established VMware's Hugging Face organization. Published 15 instruction-tuned LLMs and 2 NLP datasets, downloaded 243K+ times by researchers, engineers, and enterprises worldwide.

...
Cumulative Downloads
Across 15 open-source models and 2 datasets published under the VMware organization.
Models
Datasets

Publications

Peer-reviewed publications at ICML, EMNLP, and IEEE, plus technical articles and open-source research. View full profile on Google Scholar.

TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning 2026
Zhepei Wei, Xiao Yang, Kai Sun, Jiaqi Wang, Rulin Shao, Sean Chen, Mohammad Kachuee, Teja Gollapudi, Tony Liao, Nicolas Scheffer, Rakesh Wanga, Anuj Kumar, Yu Meng, Wen-tau Yih, Xin Luna Dong
ICML 2026

Introduces a reinforcement learning framework that directly optimizes LLM truthfulness by reducing hallucinations by 28.9% and improving truthfulness by 21.1%, while enabling models to abstain when uncertain. Currently deployed in production at Meta Reality Labs, addressing critical AI safety priorities.

PrismRAG: Boosting RAG Factuality with Distractor Resilience and Strategized Reasoning 2025
M. Kachuee, T. Gollapudi et al.
EMNLP 2025 · Industry Track

Proposes an efficient fine-tuning framework that improves RAG factuality by 5.4% by training models to handle distractor passages and plan reasoning strategies, now deployed in production systems.

The Unreasonable Effectiveness of Eccentric Automatic Prompts 2024
Rick Battle, Teja Gollapudi
arXiv 2024

Demonstrates that AI models automatically generate more effective prompts than human engineers through systematic optimization, discovering unexpected high-performing strategies (e.g., Star Trek roleplay) that boost mathematical reasoning in LLMs. Featured in IEEE Spectrum, New Scientist, and 17+ major publications, reaching 500K+ engineering professionals.

Fast Multi-scale Face Detection CNN 2019
N. S. Gollapudi, Vanitha M.
ViTECoN 2019 · IEEE

Develops an efficient CNN-based face detection model achieving 72% mAP on challenging multi-scale, multi-angle, and occluded face scenarios, running at 60fps on consumer GPUs for real-time applications.

Technical Writing

Speaking & Contributions

Peer review and program committee service at international AI conferences.

Program Committee & Peer Review
Program Committee Member
  • AIAI 2026, 22nd International Conference on AI Applications & Innovations
  • EANN 2026, 27th Engineering Applications of Neural Networks Conference
  • Additional conferences in AI safety, machine learning, and NLP
Peer Review Service

Reviewed 10+ manuscripts across top-tier venues, evaluating research on AI safety, hallucination reduction, and machine learning systems.

Coverage & Recognition

Coverage of my research and expert perspectives in leading technology publications, and more.

IEEE Spectrum May 2024
AI Prompt Engineering Is Dead
Research by Battle & Gollapudi at VMware showing AI models outperform human prompt engineers, featured in the May 2024 print issue of IEEE Spectrum.
Business Insider Mar 2024
ChatGPT and Other AI Models Can Write Better Prompts Than Humans Can
Business Insider covers VMware research showing automated prompt optimizers consistently beat human-crafted prompts across large language models.
New Scientist Feb 2024
AIs Get Better at Maths If You Tell Them to Pretend to Be in Star Trek
New Scientist covers how eccentric AI-generated prompts dramatically improve LLM mathematical reasoning, including surprising sci-fi roleplay findings.
The Register Feb 2024
Prompt Engineering Is a Task Best Left to AI Models
Coverage of research demonstrating that automatic prompt optimization consistently outperforms human-crafted prompts across Mistral and Llama 2 models.
Computerworld Mar 2024
AIs May Be Better at Prompt Optimization Than Humans
Computerworld highlights VMware research showing AI-generated prompts outperform human engineers in optimizing LLM performance on benchmark tasks.
Business Insider Feb 2024
Using Star Trek Prompts Can Boost an AI Chatbot's Basic Math Performance
Second Business Insider feature on eccentric AI-generated prompts dramatically improving LLM reasoning, including a viral Star Trek-themed example.
ITPro Mar 2024
This Engineering Discipline Was Hailed as the Next Big Thing, but AI Has Killed It Before It Even Started
ITPro covers how VMware research upends the prompt engineering profession: AI models now optimize their own instructions better than human specialists.
Government Technology 2024
Can AI-Written Prompts Yield More Accurate Math Answers?
GovTech explores the educational implications of research showing AI-generated prompts significantly improve mathematical reasoning in language models.
Reworked Mar 2024
Prompt Engineers Are Unnecessary. Long Live the Prompt Engineer
VMware research shows LLMs generate more effective prompts than humans, yet experts argue the role will evolve rather than disappear, integrating across job functions.
Synergaize Apr 2024
AI Optimized Prompts are Highly Effective and SERIOUSLY Weird
Coverage of research finding that AI-generated prompts dramatically outperform human-crafted ones, with bizarre creative narratives like Star Trek references boosting math reasoning.
Communications of the ACM 2024
Automating Tools for Prompt Engineering
ACM, the world's largest computing society, covers research demonstrating that automated prompt optimization surpasses human prompt engineers across major LLM benchmarks.
Localogy May 2024
The Death of the Prompt Engineer
Industry analysis citing VMware research as evidence that specialized prompt engineering skills are becoming obsolete as AI models learn to self-optimize instructions.
Slashdot Mar 2024
AI Prompt Engineering Is Dead
Slashdot covers research showing AI language models autonomously devise superior prompts: "In almost every case, this automatically generated prompt did better than the best prompt found through trial-and-error."
GIGAZINE Mar 2024
'Prompt engineers' who operate AI are being eliminated by AI
Japanese tech outlet GIGAZINE covers how automated prompt optimization reduces engineering time from days to hours while improving output quality, putting the profession at risk.

Get in Touch

Open to conversations about frontier AI research, whether collaborations, advisory roles, or the right next challenge.