Siddhant Garg

Software Engineer

Sunnyvale, California, United States6 yrs experience

AI EnabledHighly Stable

Key Highlights

Expert in memory-augmented LLMs for Smart Glasses.
Achieved state-of-the-art results in Open Domain Question Answering.
Presented at AAAI 2020 with impactful research.

Stackforce AI infers this person is a skilled AI Research Engineer specializing in Natural Language Processing and Machine Learning.

Contact

Skills

Core Skills

Natural Language Processing (nlp)Machine Learning

Other Skills

Memory-augmented LLMsSmart GlassesWearables GenAIPost-training LLMsOpen Domain Question AnsweringRAGAdaptive fine-tuning techniquesAuto-encoding LLMsQuestion answeringInformation retrievalDataset buildingAutomated DebuggingIntel Processor TraceControl flow informationOptimization techniques

About

I am a Research Engineer working on memory-augmented LLM experiences for Smart Glasses in the Meta Wearables GenAI team. Prior to this, I worked as an Applied Scientist II for 3 years at Amazon where I was post-training generative and extractive LLMs for Open Domain Question Answering. I graduated from the University of Wisconsin-Madison with a Masters in Computer Science in May 2020. I completed my Masters Thesis under the guidance of Prof. Yingyu Liang on representation learning paradigms and their application to text classification. Prior to this, I completed my Bachelors in Computer Science and Engineering from IIT Bombay in August 2018 where I was advised by Prof. Sunita Sarawagi.

Experience

Amazon

2 roles

Applied Scientist II

Jun 2020 – Sep 2023 · 3 yrs 3 mos · San Francisco Bay Area

Amazon Alexa: Post-training LLMs for Open Domain Question Answering + RAG driving natural, factual, contextual and conversationality improvement for Alexa Info verticals

Post-training LLMsOpen Domain Question AnsweringRAGNatural Language Processing (NLP)Machine Learning

Applied Scientist Intern

Jun 2019 – Sep 2019 · 3 mos · Manhattan Beach, CA

Adaptive fine-tuning techniques for auto-encoding LLMs in question answering and information retrieval
Contributed to build ASNQ, a large dataset based on Natural Questions (NQ) for AS. Established new state-of-the-art MAP results on WikiQA & TREC-QA with 92% & 94.3%, beating 83.4% and 87.5% from previous work. Research paper from work accepted at AAAI 2020 with Oral Presentation.

Adaptive fine-tuning techniquesAuto-encoding LLMsQuestion answeringInformation retrievalDataset buildingNatural Language Processing (NLP)+1

Epfl (école polytechnique fédérale de lausanne)

Summer Intern

May 2017 – Jul 2017 · 2 mos · Lausanne Area, Switzerland

I was a research intern under Prof. James Larus in the Very Large Scale Computing (VLSC) Lab at EPFL. I worked on the project of Automated Debugging where I worked on incorporating the new tracing tool by Intel: Intel Processor Trace (PT) into the system for automated debugging
by getting the control flow information for an execution. I along with my PhD guide Bogdan Alexandru Stoica, designed and implemented optimizations in the decoding library to reduce the overhead in tracing using an in-memory control flow graph and memoizing the already decoded instructions. We also worked on reducing the memory print by saving the decoded trace of the tail of the execution.

Automated DebuggingIntel Processor TraceControl flow informationOptimization techniques

Technical university of braunschweig

Summer Intern

May 2016 – Jul 2016 · 2 mos · Braunschweig Area, Germany

I was a research Intern under Prof. Sandor Fekete in Algorithms Group of Technische Universitat Braunschweig. I worked primarily on 'Online Algorithms for Swarm Robotics'. I read and worked on a variety of online algorithmic problems in the fields of swarm robotics and computational geometry like collective tree exploration, online square packing, etc. I worked on a new lower bound on the competitive ratio and a finding a generalized algorithm for the problem of minimizing the time of exploration of an unknown tree by a swarm of 3 robots.

Online AlgorithmsSwarm RoboticsComputational Geometry