Ashish Jaiswal

Menlo Park, CA, USA · (682) 433-6518 · ashiz2013@gmail.com

I am currently employed as a Research Scientist at Meta, where I collaborate with the Navigation team on the Facebook app. My primary focus is enhancing users' navigation experience within the app by utilizing Generative AI (GenAI) and machine learning to suggest relevant Badging and Notification signals.

I earned my Ph.D. in Computer Science in 2023, specializing in machine learning and computer vision, with a specific emphasis on improving human-computer interaction. With over four years of research experience, I excel in applying advanced data-driven methods and analytics to derive valuable insights. My expertise spans across statistical analysis, supervised and unsupervised machine learning (ML), deep learning (DL) algorithms, and signal processing. I have successfully conducted classification and regression analyses on various data modalities.

In addition to my primary focus, my research delves into self-supervised learning for effective representations of high-dimensional data and the integration of multiple data modalities to achieve superior performance. I have authored more than 15 publications, with my work being cited over 1200 times. I am dedicated to applying my knowledge and experience to drive innovation and advance the field of human-computer interaction through the application of machine learning and computer vision technologies.



Download my Resume
Download my CV

Experience

Research Scientist (Generative AI)

Meta, Menlo Park, CA
  • Facebook Navigation ML Team
  • Enhancing user navigation experience in the Facebook app by incorporating relevant badging and notification signals through machine learning and generative AI.
Jan 2024 - Present

Graduate Research/Teaching Assistant

Computer Science & Engineering, UT Arlington
  • Built pre-processing pipelines to analyze visual data for understanding cognitive fatigue in people (images, videos, fMRI)
  • Developed a multi-modal fusion algorithm to recognize Human Activities from RGB videos for cognitive analysis in children
  • Assistant Lecturer for Human-Computer Interaction (Special Topics) (5369/6369) course
  • Previously TA for programming and system courses: C, JAVA, Linux, Digital Circuits, and Raspberry Pi Labs
  • (Heracleia HCC Lab)
August 2019 - Dec 2023

Software Engineering Intern, Machine Learning (Ph.D.)

Meta, Menlo Park, CA
  • Audio Video Understanding Team
  • Built a system that visualizes the current popular trends going on in REELS videos uploaded on Facebook and Instagram
  • My POC was used to pitch in services for other clients in the company by my team manager.
  • Unsupervised Clustering of videos, NLP to assign topics to clusters generated, UI Dashboard, and Search Indexing API for video search based on topics generated.
May 2022 - August 2022

Scientific Applications Programmer

SocialEyes NP
  • Researched & implemented computer vision techniques with transfer learning to detect diseases from retinal images targetting macular degeneration
  • Built a risk analysis matrix dashboard to visualize patient's health information
April 2019 - July 2019

Software Engineer

Insight Workshop
  • Developed the core internal system of a national frozen beverage enterprise in the US inciting a high ground in its space and extending growth rate
  • Built MVP for a healthcare startup in the US that digitizes real-time health information for patients with IoT and empowers the practitioner-patient relationship
  • Developed the core internal systems of two similar tech-enabled healthcare services companies that facilitate consumer communication, engagement, and retention throughout the lifecycle of a consumer
April 2018 - May 2019

Machine Learning Engineer Trainee

Insight Workshop
  • Learned & implemented classical machine learning techniques in web applications
  • Worked on IoT projects related to health monitoring system
  • Worked with AWS Services related to compute and storage
December 2017 - April 2018

Secretary at Robotics Club

Kathmandu Engineering College
  • Worked as an acting secretary of the club
  • Worked on projects involving Robotics, Embedded Systems and programming languages like C/C++, Arduino, AVR, etc
  • Participated in different national and international robotics competitions representing club and college.
June 2015 - July 2018

Innovator

National Innovation Center
  • Programmed in C and AVR for stabilization of medical drone using LQR (Linear-Quadratic Regulator) techniques
  • Customized flight controller for accessible GPS tracking in rural places of Nepal
October 2016 - September 2017

Education

University of Texas at Arlington

Ph.D. in Computer Science
Computer Vision, Machine Learning, Human Computer Interaction
August 2019 - Dec 2023

Kathmandu Engineering College

Bachelor of Electronics & Communication Engineering
Embedded Systems, Robotics, Computer Science
November 2014 - October 2018

Pentagon International College (High School)

Physical Science
July 2011 - May 2013

Skills

Programming Languages & Tools
EXPERIENCED IN
  • Machine Learning, Computer Vision, Self-supervised Learning
  • Python, Pytorch, Numpy, Pandas, Matplotlib, Scikit-learn, Keras (TensorFlow)
  • Git, Github, Bitbucket, Gitlab
  • MySQL, AWS Aurora
  • JS (ES5, ES6), jQuery, Angular TS
  • AWS Services (EC2, RDS, S3, Lambda, etc)
  • Bash, Zsh, Linux, Nginx
  • REST API & Services (Django-REST, Flask)
  • IoT, Robotics & Embedded Systems
  • C/C++, AVR, Arduino
  • Agile Development

Interests

  • Computer Vision & Machine Learning
  • Self-supervised Learning / Semi-supervised Learning
  • Internet of Things (IoT)
  • Web Services
  • Cricket
  • Music
  • Traveling

Apart from being a machine learning enthusiast and a web developer, I enjoy most of my time in cricket. Whenever I get any chance to play indoor or outdoor cricket, I rush myself to it. I have also been an U-16 player for Kathmandu during my school days. I have participated in inter-college cricket competitions from KEC too.

When forced indoors, I follow a number of sci-fi and thriller genre movies and television shows. I spend a large amount of my free time exploring the latest technological advancements in computer science.



Checkout my past blogs (Electronic Axis)

Awards & Certifications

Projects

A Deep RNN model to predict fatigue level in subjects with/without Traumatic Brain Injury (TBI) based on their brain fMRI (3D) scans when performing certain tasks.

See Project
In the second course "Machine Learning" of the Micromasters in AI program, I was able to implement following classical machine learning techniques to carry out meaningful applications.
  • Ridge Regression & Active Learning
  • Bayes Classifier Implementation
  • Clustering (K-means & EM Gaussian)

See Project
With an intention to score high in the GRE (Graduate Record Examination), I focused on building something that could help me grasp obscure vocabularies subconsciously. So, I wrote a python script that changed my desktop wallpaper each minute with a random GRE vocabulary at a rate of 10 words per day. This project was one of the reasons why I scored 321 in the GRE.

See Project
As a part of the first course "Artificial Intelligence" from Columbia Micromasters program, I worked in the following projects.
  • Search Algorithms (BFS, DFS, A*) for 8-puzzle game
  • Adversarial Search & Game (2048 game playing agent)
  • CSP (Sudoku Solver)
  • NLP (Sentiment Analysis on IMDB reviews)

See Project
KrishiSathi is a web portal for farmers integrated with Machine Learning and IoT to analyze daily crops and their growth. The main purpose of initiating this undertaking for my "Final Year Project" was to get an insight into the integration of hardware, software and AI (learning)

See Project
This project yielded us the "Best Engineering Award" at Robocon Nepal 2018. The purpose of this robot was to make it turn-on street lights manually and then clean solar panels autonomously with the help of a helper robot. This was one of the biggest scaled robots that our team from Robotics Club had built. Here's a small video on behind the scenes while we built the robot: Youtube Video

See Project
Currency Predictor is a Django web application built to learn time series that shows the history of prices for different cryptocurrencies and also predicts their prices in the near future.

See Project
Robotic Arm powered with record-and-play feature was a college project in my junior year. The arm was Arduino driven and controlled with servo motors. The precision of each axis movement was +- 2 cm. The project was intended as a prototype for industrial chores.

See Project
With three other fellow computer scientists, we were able to build a hexacopter by customizing our own flight-controller named Hobbyking KK2. Hobbyking KK2 was intended as a prototype for the farmers that could empower them to sprinkle pesticides and monitor their crops remotely. With this project, we later participated in an international drone competition "KrishakCopter" and were the only finalist from Nepal.

See Project