Hi! I'm Sriharsha

A Software Engineer & Machine Learning Enthusiast

About Me

Hey! I am Sriharsha, currently a Graduate student in Computer Science at Manning College of Information & Computer Science at Umass Amherst. Before this, I was working as a software engineer and worked on large software systems spanning backend API development and UI development for set top box devices. My interests lie in end to end development of Machine Learning Systems, Data engineering, and Natural language processing. Additionally, my research interest lies in Natural language processing, including areas such as Long form NLP, improving evaluation metrics to effectively understand, improve current models. I did my undergrad with a major in computer science from PES University..

Outside of work, I love hiking and running (Prefers Mountains over beaches) playing and watching Tennis.

TL;DR? Self Proclamations:

Software Engineer

OSS Enthusiast

Runner

Publications

List of all the publications (Published and in-progress)

[3]Sriharsha Hatwar, Virginia Partridge, Rahul Bhargava, Fernando Bermejo

"Author Unknown : Evaluating Performance of Author Extraction Libraries on Global Online News Articles".

In Submission at AAAI-ICWSM 2024 [IN REVIEW]

[2]Aarohi Srivastava, Abhinav Rastogi, .. , Sriharsha Hatwar, .., Ziyi Wu

"Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models" [LINK].

Published at TMLR 2023 [ACCEPTED]

[1] Prajwal Chandrashekaraiah, Pranav kashyap, Sriharsha Hatwar, Srinivasa Murthy.

"Lead Artist Identification from Music Videos using Auto-encoders" [LINK]

Symposium on Machine Learning and Metaheuristics Algorithms, and Applications (SoMMA 2019) [ACCEPTED]

Technologies Known

I've been working on enhancing enterprise systems along with web development internal to the org for a couple of years now.

Some technologies I've worked with:

HTML

CSS

JQuery

Bootstrap

MongoDB

Pytorch

Keras

sklearn

Flask

PostgreS

Redis

Pandas

Languages that I have worked with:

Python

Haxe

CPP

Java

Personal Blog

List of technical, non technical and personal blogs that I have written

Personal : Kedarkantha trek

Technical : PEGASUS : Pre training with Extractive Gap sentences for Abstractive summarization

Technical : Get to the point : Summarization with Pointer-Generator Networks

Go to Site

Take me to the blog

Professional Experience

I have been doing UI and Middleware development in linux and Android based systems at the STB division in Xperi. Worked on NLP and Knowledge graph during my internship.

Xperi Corp. (Formerly TiVo)

(Nov 2018 - Present) Bangalore, India

Software Engineer

June 2019 - Present

Worked on UI Enhancement, features, and bug fixes in the core UI/UX team that drives features for the latest Tivo products. [Linux STBs and recently launched streamers]
Implemented a new design of deeplinking of assets in streamers to third-party apps using Android Intent API and eliminated the metadata team’s effort of manually creating deeplinking strings.
Redesigned an existing AWS Lambda Service to store the recent video provider and implemented a client feature to launch the provider (HBO/Prime etc.) app directly when the asset (Series/Episode) is selected for streaming
Implemented Retry mechanism for IP Linear playback for streamers following an exponential backoff policy. Built a Splunk dashboard for monitoring the devices in the field.
Redesigned TvBeacon Service for in-home device communication and reduced the round trip by 50% between mini and host STB for the latest Series 7 Set-top-boxes impacting 10k+ devices.

Associate Software Engineer

Jul 2018 - June 2019

Exposed to haxe cross-compiler toolkit, allowing one to target both Linux STBs (C++) and modern android based (Java) app for streamers.
Redesigned UI Progress bar widget across screens, worked on automation tool for sending release notes, perform cherry pick using a webbased interface.
Implemented the HPK logging for the Bluetooth based remote for enhancing debugging.
Languages and Tech Stack : (Haxe, C++, C, Python, Java - Android, JS), (Postgres, splunk, Flask, Sqlalchemy, bootstrap)

Software engineering Intern (Knowledge Graph Department)

Jul 2018 - June 2019

Worked on extracting metadata information from category section in wiki articles and enhanced metadata on media entities by around 9% overall. Also created webpage for comparing metadata results before and after KG creation
Implemented a rule based engine to mine information. Later, parallelized the creation a Knowledge graph(KG) by utilizing map-reduce pipelines.
Improved the KG by utilizing smart tags for each entity by exploiting parent-child relationship between category nodes, created the complete KG by following two pass mechanism.
Languages and Tech Stack : Python - Spacy, NLTK, Scipy, numpy; HTML, Bootstrap, JS.

Volunteering

To enhance my lacking research experience in machine learning, I have been an active member of mlcollective, a non profit org.

MLCollective

(Jun 2021 - Present)

Research member

A non-profit ML research organization, where I am working on two collaborative projects.
Working on researching robust methods for Out of distribution generalization. Additionally, working on blog post submission for ICLR 2022 in collaboration with the Natural-language-processing-subgroup
Assisted in solving a bug in MLCollective website, later provided a proper fix. [LINK].