Sravan Bodapati

AI Researcher

San Francisco, California, United States14 yrs 1 mo experience
Most Likely To SwitchHighly Stable

Key Highlights

  • Led large-scale LLM initiatives at Amazon.
  • Authored 50+ patents in AI domain.
  • Scaled teams from 2 to 30+ in 3 years.
Stackforce AI infers this person is a SaaS expert specializing in AI-driven solutions for Conversational and Natural Language Processing.

Contact

sravanbabu.bodapati@gmail.comLinkedIn

Skills

Core Skills

Large Language Models (llm)Engineering ManagementConversational AiNatural Language Processing (nlp)Speech Recognition

Other Skills

AlgorithmsAmazon Web Services (AWS)ApacheApache SparkCC++Customer ExperienceData StructuresGenerative AIGenetic AlgorithmsHadoopJavaLinuxMapReduceMatlab

About

* π‘½π’Šπ’†π’˜π’” & π‘Άπ’‘π’Šπ’π’Šπ’π’π’” 𝒂𝒓𝒆 π’•π’π’•π’‚π’π’π’š π’Žπ’Šπ’π’†, 𝒏𝒐𝒕 π’Žπ’š π’†π’Žπ’‘π’π’π’šπ’†π’“'𝒔 - Sravan currently leads and manages the Foundation Modeling Inference science efforts - spanning the problem of 1) Long Context Support 2) Speculative Decoding 3) Agentic workflows & 4) Novel LLM Modeling architectures. - Sravan previously led and managed the efforts for Medical NLP at AWS, ASR science efforts in AWS Lex, building SOTA Text Summarization, Speech to text models for AWS customers in the domain of Conversational AI for various languages. - Sravan scaled the AWS LLMs and ASR Applied Science team 2 to 30+ in 3 years spanning multiple teams across Seattle, Santa Clara, New York and Bangalore & delivered highly impactful features (like Text Summarization, Styled Slots, Custom Vocabulary, high-quality model upgrades) to the customers of AWS Lex and AWS Connect. - Sravan has onboarded several high-profile customers for AWS, and ensured a high-quality bar for their requirements. Sravan also drives cross-organizational efforts on In-context learning and prompt-tuning of LLMs (Large Language Models) - Sravan has also authored 50+ patents (pending approval from USPTO) in the domain of AI, and his research work has been published in many top tier AI conferences like ACL, KDD, ICDM. Sravan is also a reviewer of publications at ACL, EMNLP. He teaches AI/ML/NLP & Deep Learning courses at Amazon internal ML University. - Sravan was the lead scientist & manager for the development and launch of * AWS Transcribe Medical, * Custom Text classification on AWS Comprehend. * PII Content Redaction for AWS Transcribe. * LDA Topic Modeling launch on AWS SageMaker and AWS Comprehend * Ensured smooth delivery of 3 S-Team goals at Amazon, AWS; Sravan has an overall 13+ years of experience in building ML models at scale, 6+ years of technical expertise in leading/managing large science teams to deliver scientifically novel work.

Experience

Amazon

2 roles

Principal Scientist and Senior Manager of Applied Science - Amazon Nova Foundation Models

Mar 2024 – Present Β· 2 yrs Β· On-site

  • 1. π‘³π’π’π’ˆ π‘ͺ𝒐𝒏𝒕𝒆𝒙𝒕 𝒔𝒖𝒑𝒑𝒐𝒓𝒕 𝒇𝒐𝒓 𝑳𝑳𝑴𝒔 : Spearheaded technical innovation and execution for delivering 300k - 1M context support for Amazon Nova LLMs
  • 2. π‘Ίπ’‘π’†π’„π’–π’π’‚π’•π’Šπ’—π’† π‘«π’†π’„π’π’…π’Šπ’π’ˆ :Novel Algorithmic development to deliver state of the art latency and throughput to AWS customers with Amazon Nova
  • 3. π‘Ήπ’†π’‚π’”π’π’π’Šπ’π’ˆ & π‘¨π’ˆπ’†π’π’•π’Šπ’„ π’˜π’π’“π’Œπ’‡π’π’π’˜π’” : Improving the performance of Reasoning Models / Agentic workflows through Inference time scaling/Test time compute.
  • 4. 𝑰𝒏𝒇𝒆𝒓𝒆𝒏𝒄𝒆 π’‚π’˜π’‚π’“π’† π‘¨π’“π’„π’‰π’Šπ’•π’†π’„π’•π’–π’“π’†π’” : Develop LLM architectures that are more inference friendly, optimized for latency, throughput and cost!
Generative AILarge Language Models (LLM)Conversational AIEngineering ManagementTeam Management

Principal Scientist and Senior Manager of Applied Science - Amazon Rufus

Jun 2023 – Present Β· 2 yrs 9 mos Β· On-site

  • LLMs in Amazon Search!
  • Concieved and led LLM based Autocomplete, Speller and Related Searches offering on Amazon Search page.
  • Presented 2 talks at Amazon scale on 1) Utilizing LLMs for Retrieval 2) Rufus Architecture and Challenges
  • Launched Amazon RUFUS - Conversational Shopping Assistant for Amazon customers: We developed Rufus, a Generative AI powered conversational shopping assistant for Amazon. This assistant uses a custom inhouse built LLM with world knowledge, shopping knowledge based on Amazon shopping data, and personalized preferences of users. We developed it using large scale pretraining, LoRA finetuning, RAG based retrieval augmentation, ReWOO based planning, and augmenting factual responses using search and retrieval tools.
Generative AILarge Language Models (LLM)Conversational AISearch Engines

Amazon web services (aws)

5 roles

Sr. Manager of Applied Science - Large Language Models & Generative AI

Aug 2022 – Present Β· 3 yrs 7 mos

  • Heading Large Language Models initiatives for Summarization, Faithfulness and other NLP tasks in the realm of Generative AI
Engineering ManagementNatural Language Processing (NLP)Large Language Models (LLM)

Head / Sr.Manager of Applied Science - Lex ASR

Promoted

Aug 2021 – Present Β· 4 yrs 7 mos

  • Head of all the ASR efforts for AWS Lex; My team works on building base ASR models and contextualizing the models for Conversational AI/ Goal Oriented Dialog use-cases.
  • I lead the OP1/scientific, technical, product roadmap and own the technical initiatives charter for AWS Lex ASR to invest in, for the upcoming years.
  • I also drive cross-organizational efforts on In-context learning and prompt-tuning of LLMs (Large Language Models)
Engineering ManagementSpeech RecognitionNatural Language Processing (NLP)

Applied Science Manager

Jul 2019 – Mar 2023 Β· 3 yrs 8 mos

  • Leading automatic speech recognition(ASR) efforts for Conversational AI offering of AWS.
  • Specialized in Language Modeling/NLP/Personalization.
  • https://aws.amazon.com/lex/
  • Accomplishments:
  • 1. Scaled the team from 2 scientists to a very high-performing team of 20+ scientists
  • 2. Published state-of-the-art research at top tier conferences in ASR and NLP through my team
Engineering ManagementSpeech RecognitionNatural Language Processing (NLP)

Senior Applied Scientist

Jan 2019 – Apr 2020 Β· 1 yr 3 mos

  • I'm currently working as Senior Scientist in AWS Transcribe, building SOTA Speech to text models for AWS customers for various languages.
  • I'm the lead scientist for the launch of Amazon Transcribe Medical : https://aws.amazon.com/transcribe/medical/
  • I'm the lead scientist for the launch of Content Redaction (PII entity redaction) on AWS Transcribe:https://docs.aws.amazon.com/transcribe/latest/dg/content-redaction.html
  • I have helped onboard few high profile customers for AWS, and ensured quality bar for their requirements.
Speech RecognitionCustomer Experience

Applied Scientist Lead, AWS Comprehend

Jan 2017 – Dec 2018 Β· 1 yr 11 mos

  • Deep Learning at AWS AI.
  • Lead scientist to develop and launch Custom Text classification on AWS Comprehend.
  • Worked on high profile customer usecases in the domain of NLP and hit customer goals on them.
  • Delivered LDA Topic Modeling on AWS SageMaker and AWS Comprehend

Wecnlp2020

Chair - Member of Organizing Committee

Apr 2020 – Oct 2020 Β· 6 mos Β· San Francisco Bay Area

  • https://www.wecnlp.ai/
  • 3rd Edition of this Annual Conference for collaboration with top NLP researchers in academia and industry.

Amazon

ML & NLP Research Engineer II / Applied Scientist II

Jan 2013 – Apr 2017 Β· 4 yrs 3 mos Β· Greater Seattle Area

  • Language Detection for eBook Content
  • Auto identification of Asins that belong to same series/bundle. For ex : Given HP Sorcers stone, auto identification of other asins in HP series
  • ML/NLP Techniques for Auto identification of Start Reading Location (SRL) of an eBook
  • ML/NLP classifiers for Content Mismatch Identification with the metadata of an eBook
  • Customer lifecycle modeling to identify churn propensity of customers, and identifying right mechanisms to target them to prevent churn, increase activation.
  • SparkML for modeling at scale.
  • Developed and shipped pricing logic for e-Book bundles, which boosted revenues in the order of millions.
  • Built Infrastructure for various pricing models like StrikeThru Pricing for Kindle.
  • Developed and productionized training pipeline using AWS services like EMR, SWF that helps in faster iterations for improving performance of models

Qualcomm

Internee

May 2012 – Jul 2012 Β· 2 mos Β· Hyderabad Area, India

  • Project Overview :
  • Python SUDS library to develop JIRA framework using SOAP and REST protocols by making RPC connections to JIRA Server
  • Details :
  • Developed Search , Push and Pull framework for JIRA and PRISM ( Internal Bug Tracking Tool of Qualcomm)

Global analytics

R & D SDE

Jan 2012 – Jan 2013 Β· 1 yr Β· India

  • Was a part of R&D Team that develops models for the lending business of Global Analytics.
  • Developed Feature Extraction Module that pre-processes the data into bins, assigns flags, performs imputations and few other transformations which specially boosts performance of Linear Classifiers

Airbus

Java Software Developer

May 2011 – Jul 2011 Β· 2 mos Β· Bengaluru Area, India

  • Hadoop Map Reduce : Invariant feature detection using Apache Mahout
  • Clustering of Vectors (which image processing libs extract from imagees )using Apache Mahout into various clusters . Algorithms for clustering are provided by Apache Mahout which runs on hadoop using the map/reduce paradigm.

Softeon

Algorithm Engineer

May 2010 – Jul 2010 Β· 2 mos Β· Chennai Area, India

Education

Indian Institute of Technology, Madras

Dual Degree ( B.Tech + M.Tech ) β€” Computer Science

Indian Institute of Technology, Madras

Computer Science

Stackforce found 100+ more professionals with Large Language Models (llm) & Engineering Management

Explore similar profiles based on matching skills and experience

Anjali Singh

Anjali Singh

Senior Technical Project Manager

at Concentrix

Delhi, India11 yrs 1 mo exp
Artificial Intelligence (AI)Project Management
VR

Venkatesh Ramaswamy

VP of Engineering

at Prosper Marketplace

San Jose, United States22 yrs 6 mos exp
Cloud ComputingGenerative AIMachine LearningSoftware Engineering
S.

Shivam .

Senior Member of Technical Staff

at Salesforce

Bengaluru, India7 yrs exp
Artificial Intelligence (AI)Backend DevelopmentCloud ComputingMachine LearningSoftware Development
Apoorva Chourasiya

Apoorva Chourasiya

Solutions Architect

at POS NIRVANA PRIVATE LIMITED

Pune, India14 yrs 3 mos exp
AI & Machine LearningCloud InfrastructureJava & Spring Boot Development
Kriti Kohli

Kriti Kohli

Senior Manager, GenAI

at Google

New York City, United States16 yrs 11 mos exp
Artificial Intelligence (AI)Machine LearningData Science
Omkar P.

Omkar P.

Chief Executive Officer

at QAI

Berlin, Germany12 yrs 8 mos exp
Artificial Intelligence (AI)Mixed RealityProject ManagementSoftware as a Service (SaaS)Team Management
Hitesh Bandhu

Hitesh Bandhu

Co-Founder

at Social Cursor

Chandigarh, India3 yrs 1 mo exp
AI EngineeringBackend Development
Prashant Sahu

Prashant Sahu

Associate Director

at Optum

Hyderabad, India14 yrs 6 mos exp
Cloud Architecture & StrategyEnterprise Digital TransformationProduct InnovationSolution Architecture