Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I'm Naga Penugonda, an AI Scientist and Machine Learning Engineer with around five years of experience in large language models, transformer architectures, multi-agent systems, and generative AI. I enjoy turning complex research into practical, cloud-native solutions that scale in both academic and enterprise settings. Most recently at Google, I built multi-agent reasoning systems and retrieval-augmented generation pipelines over 40M+ biomedical entries, reducing experimental search time and delivering interactive, context-aware NLP tools for researchers. I also led deployment of distributed training, model monitoring, and responsible AI practices on GCP, achieving measurable gains in accuracy, efficiency, and collaboration across teams.…Hi, I'm Naga Penugonda, an AI Scientist and Machine Learning Engineer with around five years of experience in large language models, transformer architectures, multi-agent systems, and generative AI. I enjoy turning complex research into practical, cloud-native solutions that scale in both academic and enterprise settings. Most recently at Google, I built multi-agent reasoning systems and retrieval-augmented generation pipelines over 40M+ biomedical entries, reducing experimental search time and delivering interactive, context-aware NLP tools for researchers. I also led deployment of distributed training, model monitoring, and responsible AI practices on GCP, achieving measurable gains in accuracy, efficiency, and collaboration across teams.

Naga Penugonda

Data Scientist, AI Engineer, AI Developer, +2





Hi, I'm Naga Penugonda, an AI Scientist and Machine Learning Engineer with around five years of experience in large language models, transformer architectures, multi-agent systems, and generative AI. I enjoy turning complex research into practical, cloud-native solutions that scale in both academic and enterprise settings. Most recently at Google, I built multi-agent reasoning systems and retrieval-augmented generation pipelines over 40M+ biomedical entries, reducing experimental search time and delivering interactive, context-aware NLP tools for researchers. I also led deployment of distributed training, model monitoring, and responsible AI practices on GCP, achieving measurable gains in accuracy, efficiency, and collaboration across teams.…Hi, I'm Naga Penugonda, an AI Scientist and Machine Learning Engineer with around five years of experience in large language models, transformer architectures, multi-agent systems, and generative AI. I enjoy turning complex research into practical, cloud-native solutions that scale in both academic and enterprise settings. Most recently at Google, I built multi-agent reasoning systems and retrieval-augmented generation pipelines over 40M+ biomedical entries, reducing experimental search time and delivering interactive, context-aware NLP tools for researchers. I also led deployment of distributed training, model monitoring, and responsible AI practices on GCP, achieving measurable gains in accuracy, efficiency, and collaboration across teams.

Available to hire

Hi, I’m Naga Penugonda, an AI Scientist and Machine Learning Engineer with around five years of experience in large language models, transformer architectures, multi-agent systems, and generative AI. I enjoy turning complex research into practical, cloud-native solutions that scale in both academic and enterprise settings.

Most recently at Google, I built multi-agent reasoning systems and retrieval-augmented generation pipelines over 40M+ biomedical entries, reducing experimental search time and delivering interactive, context-aware NLP tools for researchers. I also led deployment of distributed training, model monitoring, and responsible AI practices on GCP, achieving measurable gains in accuracy, efficiency, and collaboration across teams.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

Applied AI Scientist at Google

September 1, 2024 - Present

Led a multi-agent AI system using JAX and TensorFlow to generate and validate biomedical hypotheses, integrating retrieval-augmented generation over 40M+ genomic and chemical entries; reduced experimental search time from 48 to 31 hours per batch. Trained transformer-based models for protein interactions and drug repurposing, achieving a 22% accuracy improvement on benchmark datasets. Built a cloud-native architecture on GCP with Kubernetes for scalable training and inference, cutting end-to-end deployment time by 40%. Implemented SQL pipelines for rapid retrieval of multi-modal data; operationalized NLP-driven biomedical chatbots to accelerate literature search and experimental planning by 30%. Fine-tuned transformers with Hugging Face for biomedical text understanding, enabling automated literature summarization and hypothesis extraction with 95% semantic accuracy across 10K+ articles. Conducted end-to-end R&D on multi-agent reasoning to simulate biological interactions, contributing

Machine Learning Engineer at HCL

February 1, 2021 - August 1, 2023

Developed predictive models for 5G network traffic and customer churn with Python and ensemble methods, achieving 92% accuracy and reducing downtime forecast errors by 28%. Built BERT-based NLP pipelines to analyze telecom interactions, leveraging LLM embeddings to classify service requests and improve automated ticket resolution by 35%. Implemented end-to-end CI/CD pipelines using Docker and AWS SageMaker, reducing model release cycle time from 14 days to 5 days. Created real-time analytics pipelines on AWS that process 2M+ events daily to optimize network utilization by 18%. Performed hyperparameter optimization with Optuna and Ray Tune, boosting F1 scores by 12% and ensuring robustness. Integrated responsible AI practices and comprehensive model validation frameworks for production reliability, including drift detection and reproducibility.