Aman Mehta

AI Researcher

San Francisco, California, United States8 yrs 11 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Expert in AI and Machine Learning with proven research impact.
Developed frameworks that significantly reduced engineering hours.
Mentored projects enhancing neural models for question-answering.

Stackforce AI infers this person is a highly skilled AI researcher and software engineer with a focus on machine learning and data processing.

Contact

Skills

Core Skills

Machine LearningArtificial Intelligence (ai)

Other Skills

Deep LearningDistributed SystemsStorageDatabasesCacheData StructuresNode.jsCC++PythonShell ScriptingProgrammingObject-Oriented Programming (OOP)Agile MethodologiesSoftware Design

Experience

8 yrs 11 mos

Total Experience

1 yr 6 mos

Average Tenure

1 yr 9 mos

Current Experience

Snowflake

Senior Researcher

Sep 2024 – Present · 1 yr 9 mos · San Francisco Bay Area · Hybrid

Agents:
leading continual learning (memory)
reasoning and planning
eval, llm-judge
Search:
instruction-following retrieval
quality, eval and autotune

Machine LearningDeep LearningArtificial Intelligence (AI)Distributed Systems

X

2 roles

Software Engineer

Oct 2022 – Nov 2024 · 2 yrs 1 mo

Software Engineer

Jun 2021 – Sep 2022 · 1 yr 3 mos

Motorq

Software Engineer

Jun 2020 – Jan 2021 · 7 mos · India · Remote

Google summer of code

Mentor

May 2019 – Aug 2019 · 3 mos · London, United Kingdom

Co-founded and mentored a project on improving DBpedia's neural Question-Answering model. The project aims specifically at improving the model for compositional questions.

International institute of information technology hyderabad (iiith)

3 roles

Teaching Assistant

Jan 2019 – Nov 2019 · 10 mos

1) Computer Programming
2) Data Structures and Algorithms
3) Databases

System And Database Administrator

Aug 2017 – Jul 2018 · 11 mos

Researcher under Dr. Kamal Karlapalem

May 2017 – May 2021 · 4 yrs

Master's thesis defended on Text Driven Knowledge Graphs.
Received AI specialization certificate.
State of the art Knowledge Graph construction; beating T2KG in quality and cost; F1 score of 67% and 10^7 less computation cost
Designed and developed a hierarchical multi-stage Siamese Network model for Predicate Mapping
Developed SimX - KG based document similarity classifier along with providing explanation for the classification result
Publication - Scalable Knowledge Graph Construction over Text using Deep Learning based Predicate Mapping - https://dl.acm.org/doi/10.1145/3308560.3317708
At The Web Conference 2019, SF

Machine LearningDeep LearningArtificial Intelligence (AI)

Google summer of code

Google Summer of Code with DBpedia association

May 2018 – Aug 2018 · 3 mos

Working to expand DBPedia's neural machine learning model for automatic Question-Answering.
Extended the domain of DBpedia’s Neural Question-Answering
(QA) model by 100x and improved quality of answers by 80%
Developed a heuristic to generate training dataset; pre-processing
and post-cleaning to achieve higher model performance
Co-founded and mentored a project to expand DBpedia’s Neural Network model for compositional and complex QA