Sriram Mahesh

Bay Area · California ·
(+1) 669-649-3768 ·
(+91) 9444929178 ·
sriramja@gmail.com | srmahesh@ucsc.edu

Student Ambassador and a Masters student in Natural Language Processing (Major: Artificial Intelligence) at the University of California, Santa Cruz (UCSC).

3+ years of work experience in AI & Data Science at an MNC.

Education

University of California at Santa Cruz

Master of Science (M.S.) in Natural Language Processing
[ MAJOR: ARTIFICIAL INTELLIGENCE ]

Program Student Ambassador

Coursework: Machine Learning, Linguistic Models, Data Collection, Wrangling, Natural Language Processing (I, II, III), Advanced Machine Learning, Industry Seminars, Capstone Project. (Course Curriculum)

CGPA: 3.9 / 4.0 (To date)

2021 - 2022

SRM Institute of Science and Technology

Bachelor of Technology (B. Tech.) in Computer Science and Engineering

Coursework: Artificial Intelligence & Expert Systems, Cloud Computing, Data Structures & Algorithm Design, Web Technology, Database Management Systems, Software Engineering, Object Oriented Programming, Linux Internals, Network Security, Ethical Hacking

CGPA: 8.4/10.0

2014 - 2018

Work Experience

Software Engineer (Role: AI Data Scientist)

Larsen & Toubro Infotech - LTI

ROLE: Work on R&D projects for Fortune 500 Clients to develop solution accelerators and Minimum Viable Products (MVPs) as part of Intellectual Property (IP) Initiatives.

Banking & Financial Services (BFS) Domain

Development of a Knowledge Search Engine using NLP & Elasticsearch - Application parses through the Local corpus (PDF/Word in Single/Batch) - Both Machine Readable & Scanned (using OCR) and structures the information (Text, table, images etc) in a React application. Solution also crawls information from multiple web sources and displays relevant answers. Deployed application in Microsoft Azure.
Information Extraction in Legal domain and Depict Insights - Applied OCR & built an NLP ontology model for information extraction and context understanding from a huge volume of financial contracts using Python, APIs and IBM Cloud Watson Services. Depicted KPIs in Power BI and build a React UI to visualize insights.
Objective: A solution to assist clients with LIBOR Transition.
Automation of Identification and AI-based Visual Testing of UI components - Analyzes the DOM of webpage(s) and describes UI components and its styling/clientside validation properties in natural language. Also detects visual changes in UI using Computer Vision as applications undergo development over time.
Develop and Automate Corporate Data Governance Dashboards - Developed multiple Power BI dashboards & automated the refresh process for a data governance initiative. Built Key Performance Indicators (KPIs) for resource onboarding, allocation, billing & movement for the entire Business Unit.

Additional Efforts

Key Organizer for an Internal Global Hackathon (Organized by LTI and Powered by IBM ) - Coordinated with IBM (Hackathon Cloud partner) and 45+ teams (375+ participants) to address 13 problem areas faced in BFS (Business Unit).
Won 2nd Runner in an organization-level Hackathon - Proposed a solution to drastically track and control carbon emissions and also suggested to provide carbon advisory solutions to potential clients.
LTI Solvathon 2021 - SME and Technical Mentor of two key problem statements for teams in an organization-wide Hackathon -

Oil & Gas - MFG Domain

Refactor and Migrate 8 applications (Dotnet + SQL DB + ETLs) from On-Prem to Azure Cloud (PaaS) solution - Silo & Integrated Applications
Develop and deploy ETLs in SSIS, Azure Data Factory to transform and clean data from multiple sources (SQL / Oracle / Flat files)
Automate approval process, weekly status email, dashboard refresh using Microsoft Power Automate and SharePoint.

August 2018 - August 2021

CONFERENCE SPEAKER

Applications and Challenges of AI in Computer Vision and Language

IBM INTERNATIONAL DIGITAL DEVELOPER CONFERENCE 2020 - INDIA & ASIA PACIFIC

About: I was invited by IBM as a speaker at their International Digital Developer Conference 2020, along with other key speakers like Mukesh Jain (CTIO, Cognizant) and Senior Data Scientists from IBM, where I presented my thoughts and experiences on the topic “Applications and Challenges of AI in Vision and Language”.

During the technical session, I covered topics in Computer Vision and Natural language processing. In the session, I explained concepts behind some of the applications, showed execution of program codes while explaining the benefits of transfer learning, pointed out some of the challenges in the subject areas and demonstrated the benefits of performing inter-disciplinary study in the subjects to address some of the challenges.

On the LIVE session, I also got to connect with other developers and address few of their queries.

Read or download the PPT and view the session recording from the below links:

( My part in the session starts at 02:16:00 in the recording )

November 2020

CONFERENCE PUBLICATION

Depicting User Connections in Social Media by Mapping User-Shared Images

IEEE XPLORE INTERNATIONAL DIGITAL LIBRARY

About: Presented in 2018 Fourth International Conference on Computing Communication Control and Automation (ICCUBEA).

The research work involved building a recommendation engine using Computer Vision techniques and advanced Machine learning algorithms deployed on Amazon Cloud, to help determine user relations in Social Media ( Facebook, Pinterest ) based on the public images users share which involved analyzing human behavioral patterns on social networking sites based on factors like demographics and interests. The solution improves on the existing system by understanding media content shared to provide user recommendation to one another (and/or) allow businesses to find their target customers easily.
The research work was done under the mentorship of my then-head professor, Dr.S.Prasanna Devi

CONFERENCE: International Conference on Computing Communication Control and Automation (ICCUBEA)

You can download the paper or read it online from the links below:

June 2020

Skills

Programming:	Python, R, SQL, PL/SQL ,C
Frameworks	RESTful APIs, Zookeeper, Flask
Machine Learning Libraries:	Scikit-learn, TensorFlow, Keras, NumPy, Pandas, Matplotlib, OpenCV, NLTK, AutoSklearn
Operating System:	Windows, Linux
Other Technologies:	git, Oracle, MySQL, Tableau, Power BI

Projects

Analyzing Neighborhoods For Building A Mall

Use Data Science methods to determine best place to build a mall

INFOSEARCH

Information Extraction & Search from Local Corpora

Multi-Class Object Detection

Custom Object Detection Using YOLO v4 In Real-Time

First Order GAN

Using Generative Adversarial Networks to attempt Deepfake

(Research / Project)

Unsupervised Intent Detection and Question Answering System

Chatbot for Railway ticket reservation

Built a Question Answering System using NLP & Deep learning