Anshul Tomar

Product Engineer

Austin, Texas, United States2 yrs 11 mos experience

Most Likely To Switch

Key Highlights

Expert in Machine Learning and Data Science.
Proven track record in developing innovative AI models.
Strong foundation in quantitative analysis and statistics.

Stackforce AI infers this person is a Machine Learning and Data Science expert in the AdTech and Robotics industries.

Contact

anshulmanas@gmail.com LinkedIn

Skills

Core Skills

Machine LearningData ScienceDeep Learning

Other Skills

AlgorithmsAudio Visual (AV) SystemsC#C++EngineeringEnglishMicrosoft AzureNavigationPyTorchPythonR (Programming Language)ResearchRoboticsSQLSoftware Development

About

I'm currently pursuing my master's in computer science from UT Austin. Prior to this, I worked as a Data And Applied Scientist at Microsoft and graduated from IIT Bombay with a major in Computer Science and a minor in Applied Statistics and Informatics. I am deeply interested in ML/AI, Data Science and also in quantitative trading. I have a background in Machine Learning, Mathematics(INMO Awardee), and Statistics.

Experience

Parallel web systems inc.

Member of Technical Staff

Jul 2024 – Present · 1 yr 8 mos · San Francisco Bay Area · On-site

Building something new

The university of texas at austin

Graduate Research Assistant

May 2023 – Jul 2024 · 1 yr 2 mos · Austin, Texas, United States · On-site

Sim2Real transfer for audio-visual navigation under Prof. Kristen Grauman.
Proposed the first treatment of sim2real for audio-visual navigation by disentangling it into
acoustic field prediction (AFP) and waypoint navigation.
Designed a strategy that intelligently selects the best frequency band for prediction based on both the measured distribution and the energy distribution of the received audio, which improves the performance of the real data
Built a real robot platform and showed that the transferred policy can successfully navigate to sounding objects.

Microsoft

2 roles

Data and Applied Scientist

May 2021 – Aug 2022 · 1 yr 3 mos · Bengaluru, Karnataka, India

Worked in the BingAds Ads Trust and Safety team to develop image and multi-modal models for checking policy compliance of ads.
Worked with the AdsBrain team to achieve faster inference speeds for models in production

Data Scientist Intern

May 2020 – Jul 2020 · 2 mos · Bengaluru, Karnataka, India

Implemented the parallelized graphSAGE model for Large scale ads-query-user dataset using external sort and reservoir sampling which reduced training time from 4 min for the base model to 0.6 min/100 epochs.
Integrated FastText embeddings of nodes with the graphSAGE model resulting in a 11.29% gain in accuracy.
Increased AUC for click-through rate by augmenting the dataset with pseudo-counted pairs by 1.21%

National university of singapore

Summer Research Intern

May 2019 – Jul 2019 · 2 mos · Remote Internship

An attempt to improve the existing hashing methods present for model counting like the ApproxMC2. Approached the problem by trying to make the lower and upper bounds of edge isoperimetric inequalities even stronger.