Hi,

I am Yash Bitla

  • Computer Science
  • Software Engineer
  • Fullstack Engineer
  • ML / AI Engineer

Los Angeles, California

Resume

About

Hi, I am Yash Bitla

I hold a Master's degree in Computer Science from the University Of Southern California and am passionate about the intersection of Machine Learning and Software Engineering. With over two years of professional and internship experience, I am actively seeking full-time opportunities in the fields of Software Eningeering and Machine Learning starting June 2024.

Currently, I am interning at Dragonfruit AI, where I am spearheading the integration of object detection training workflows with Weights and Biases API. This role involves automating and scaling AI model training, as well as developing Docker containers to streamline setup and deployment processes, enhancing operational efficiency. Previously, at Galileo Financial Technologies (SoFi), I enhanced fraud detection capabilities by implementing a configurable name mismatch detection feature for ACH transactions, significantly improving platform security. At Development Monitors LLC, as a Machine Learning Engineer, I developed machine learning models for community mapping software, effectively bridging theoretical concepts with practical applications.

I also conduct research at USC's Integrated Media Systems Center Lab under Prof. Seon Kim, focusing on computer vision techniques to detect and count homeless encampments in Los Angeles, using real-world street-level video data.

When I'm not at my desk I am probably watching anime, playing soccer, or at a coffee shop :)

Image 1
Image 2
Image 3
Image 4

Experience

Software Engineer Intern

Dragonfruit AI

March 2024 - Present

  • Leading integration of ML training workflows with weights and biases API, improving efficiency by 5% to manage large-scale AI model training more effectively.
  • Dockerized training environments and automated model training launches, reducing deployment time by 10%.
  • Enhancing the notification alert system to periodically collect and analyze false positive alerts using Celery and RabbitMQ.
  • Implementing RESTful APIs for dynamic alert processing with client-specific configurations managed in PostgreSQL.

Software Engineer Intern

SoFi (Galileo Finacial Technologies)

May 2023 - August 2023

  • Designed and developed a robust monitoring system using Kubernetes and RabbitMQ to capture metrics and results for the fraud detection service.
  • Optimized system to run asynchronously with fraud detection service, ensuring no impact on response times.
  • Built and deployed real-time analytics dashboard using Splunk and DynamoDB, resulting in a 10% increase in client attraction and retention.
  • Gained proficiency in AWS (DynamoDB, KMS, IAM, Terraform, EKS), Octopus, and Kubernetes, successfully deploying summer project work for client validation and efficiently managing debugging processes.

Research Assistant

Integrated Media Systems Center Lab, USC

May 2022 - May 2023

  • Worked on a joint project with the LA Sanitation Department to explore and address real-world street monitoring complexities, specifically focusing on homeless encampments in Los Angeles.
  • Authored a case study on object detection and counting challenges, supervised by Dr. Seon Kim
  • Adapted the YOLOv5 model to detect homeless encampments in an occluded, illuminated etc noisy environments with a precision of 0.75 and counting error ratio of 0.13.
  • Implemented CSRT, KCF, and MIL tracker for tracking multiple objects for a non-static camera.

Software Engineer

Development Monitors LLC, Arlington, USA

January 2021 - December 2021

  • Developed and fine-tuned a Feature Pyramid Network on geospatial satellite imagery, enhancing accuracy in detecting roof types (mud, metal, clay) from an mIOU of 0.58 to 0.74.
  • Engineered an end-to-end inference pipeline, developing robust RESTful APIs to invoke ML models for efficient detection in user-selected regions.
  • Analysed API responses to dynamically generate shapefiles, transforming pixel coordinates to latitudelongitude data for precise mapping; reducing mapping error by 12%.
  • Optimized web-based rendering of detection output, leading to 2x increase in overall time efficiency.

Selected Publications

Object Detection and Counting Challenges in Real Street Monitoring: Case Study of Homeless Encampments (Link)

Abdullah Alfarrarjeh; Seon Ho Kim; Utkarsh Baranwal; Yash Bitla

2023 IEEE International Conference on Image Processing (ICIP)

Wide area urban street monitoring is highly demanding in various smart city applications. Manual monitoring is both laborious and time-consuming, hence automatic vision-based monitoring is a more feasible alternative. An essential part of vision-based street monitoring is detecting and counting objects of interest. However, these tasks are not straightforward due to various challenges, i.e., noisy conditions in a real environment, such as occlusion and high illumination. This study investigates the impact of these challenges on object detection and counting accuracy, then provides an empirical study to address the challenges with respect to video-based street monitoring. The selected case study demonstrates detecting and counting of homeless encampments in Los Angeles streets using street-level videos collected from a moving vehicle.

Education

University of Southern California, Los Angeles

Master's in Computer Science

January 2022 - December 2023

GPA - 3.61/4


Activities:
  • Course Producer for DSCI 550
  • Researcher at Integrated Media Systems Center (IMSC) Lab
  • Data Analyst at Facilities Planning & Management Services - Energy Services, Utility Billing Data Analysis

Courses:

Analysis of Algorithm, Introduction to Artificial Intelligence, Database and Management Systems, Machine Learning for Data Science, Web Technologies, Information Retrieval and Web Search Engines, Natural Language Processing

University of Mumbai, India

Bachelor's in Engineering, Computer Engineering

August 2017 - January 2021

GPA - 3.9/4


Activities:
  • Second Place at Game of Codes 2019
  • Runners Up at Smart India Hackathon 2020
  • Head of the Creative Club and Décor team in Don Bosco Institute of Technology (2019-20)

Courses:

Object Oriented Programming, Data Structures and Algorithms, Database and Management Systems, Artificial Intelligence, Machine Learning, Computer Networks

Skills

Technical Skills

As a computer science enthusiast, I have honed my expertise in languages like Python, Java, and C/C++, and have hands-on experience with data structures, algorithms, web development, and machine learning.

PYTHON
95%
HTML/CSS
85%
NODEJS
65%
MYSQL
85%
Java
Flask
Pytorch
Docker
kubernetes
ReactJS
AngularJS
RestAPI
Git
MongoDb
hadoop
AWS
Tensorflow
GCP
Keras
php
firebase
C++
linux
postgres

Projects