Summary
Overview
Work History
Education
Skills
Publications
Patents
Timeline
Generic

Ravi Choudhary

Bellevue,USA

Summary

Highly accomplished Machine Learning Engineer with 13 years of experience in software development and 8+ years of expertise in applying advanced algorithms within Machine Learning and Natural Language Processing (NLP). Proven ability to lead impactful initiatives, currently at Meta, focusing on fine-tuning large language models (LLMs), optimizing ad ranking and retrieval, and developing user-centric ML solutions that enhance communication and engagement. Passionate about leveraging cutting-edge technologies to drive innovation and deliver significant business impact.

Overview

14
14
years of professional experience

Work History

Senior Machine Learning Engineer

Meta
09.2023 - Current
  • Spearheaded ML model development to identify the top 500 ads from 30-40 million candidates within a stringent 300ms latency requirement, significantly improving ad serving efficiency
  • Improved personalized ad retrieval and ranking for Meta's diverse platforms, leading to increased user engagement and ad revenue
  • Revamped multi-stage ranking ML systems, boosting overall efficiency and effectiveness of ad delivery
  • Implemented global optimization strategies for ad candidate generation, resulting in a more diverse and relevant ad pool
  • Leveraged new hardware to enhance model training and operational efficiency, reducing training time and infrastructure costs
  • Established robust measurement and development processes to streamline workflows and improve team productivity

Staff Machine Learning Engineer

Qualcomm AI Research
07.2017 - 07.2023
  • Developed and fine-tuned Transformer-based LLMs for advanced text generation and context-driven systems, pushing the boundaries of natural language understanding
  • Built large-scale human preference datasets to train customized generative AI models, improving the accuracy and relevance of generated content
  • Improved NLP model performance using advanced techniques such as active learning and transfer learning, leading to significant gains in model accuracy and efficiency
  • Designed multi-modal audio captioning and ONNX models, achieving a remarkable 94% reduction in inference time, enabling real-time applications
  • Mentored a team to enhance quality in abstractive summarization and dialogue systems, fostering knowledge sharing and team growth

Machine Learning Co-op

Signify (Philips Lighting Research)
05.2016 - 12.2016
  • Built sentiment analysis models for lighting impact studies using social media data, providing valuable insights into public perception of urban lighting
  • Designed an interactive mapping tool to assess urban lighting’s societal effects, enabling data-driven decision-making for urban planning
  • Applied computer vision techniques for object identification in complex datasets, improving the accuracy and efficiency of data analysis

Senior Software Engineer

Honeywell Inc.
09.2010 - 08.2015
  • Designed inter-communication systems for flight management software using FSM models, ensuring reliable and efficient communication between critical aircraft systems
  • Conducted root cause analysis, achieving zero defect density in critical software systems, demonstrating a commitment to high-quality software development
  • Maintained fault tolerance and redundancy management systems for aviation software, ensuring the safety and reliability of aircraft operations

Education

Master of Science - Computer Software Engineering, Specialization in Machine Learning

University of Massachusetts
Amherst, MA
05-2017

Bachelor’s - Electronics & Communication

University of Rajasthan
06-2009

Skills

  • Python
  • Java
  • C
  • PyTorch
  • TensorFlow
  • Active Learning
  • Transfer Learning
  • SQL
  • NoSQL
  • GraphDB
  • MongoDB
  • Redis
  • ONNX
  • SNPE
  • TFLite
  • AWS
  • Docker
  • Kubernetes
  • CI/CD Pipelines
  • Deep learning
  • Machine learning
  • Complex Problem-solving
  • Interpersonal skills
  • Continuous improvement
  • Algorithm development
  • Large dataset management

Publications

Activity Report Analysis with Automatic Single or Multi-span Answer Extraction, arXiv, R Choudhary, AK Sridhar, E Visser

Patents

  • Audio analytics for natural language processing, E Visser, F Saki, Y Guo, M Sunkuk, LH Kim, R Choudhary
  • Multi-modal user interface, R Choudhary, LH Kim, M Sunkuk, Y Guo, F Saki, E Visser
  • Activity query response system, E Visser, R Mahfuz, R Choudhary, LH Kim, M Sunkuk, Y Guo, F Saki
  • Semantically-augmented context representation generation, AK Sridhar, R Choudhary, LH Kim, E Visser
  • System and method to view occupant status and manage devices of building, R Choudhary, Y Guo, F Saki, E Visser
  • User experience evaluation, LH Kim, Y Guo, R Choudhary, M Sunkuk, E Visser, F Saki

Timeline

Senior Machine Learning Engineer

Meta
09.2023 - Current

Staff Machine Learning Engineer

Qualcomm AI Research
07.2017 - 07.2023

Machine Learning Co-op

Signify (Philips Lighting Research)
05.2016 - 12.2016

Senior Software Engineer

Honeywell Inc.
09.2010 - 08.2015

Master of Science - Computer Software Engineering, Specialization in Machine Learning

University of Massachusetts

Bachelor’s - Electronics & Communication

University of Rajasthan

Similar Profiles

CREATE PROFILE
Ravi Choudhary