Hi, my name is

Sashankh.

I turn ideas into code.

A passionate full-stack developer and Machine Learning Engineer. I build APIs, web apps and data pipelines.

About Me

I am a software engineer with a passion for applied Machine Learning and distributed systems, with a BS in Computer Science. I used to be a senior engineer at VISA Singapore before I became at graduate student at Georgia Tech.
I am a full-stack developer - I build web application frontends, APIs, batch jobs and data pipelines. I am currently working at the intersection of systems and applied Machine Learning. Here are a few technologies I've been working with recently:
  • Scala
  • Java
  • JavaScript
  • Angular
  • Spring Boot
  • Apache Spark

Experience

Machine Learning Engineer Intern - Kaliber Labs Inc
May 2024 - Aug 2024

Full-stack engineering and ML Engineering at Kaliber Labs, Santa Clara.

  1. Built a state-of-the-art annotation platform for medical video data, enabling powerful annotations to train Kaliber’s core vision models.
  2. Sped-up inference of ensemble models (ONNX) with parallelization
  3. Built a new multi-modal chat application from scratch for a sophisticated question-answering experience.
Senior Software Engineer - VISA
Oct 2020 - Jul 2023

Full-stack software engineer on audience targeting platform to create marketing audiences from 3.9B cards.

  1. Designed and implemented a drag-and-drop audience builder UI using Angular 14. Users reported improved usability, responsiveness, and multi-tasking capabilities. UI supported creation of complex queries with intuitive visualizations and query validation.
  2. Proposed and led platform observability project to export operational metrics to Prometheus and Grafana and enable alerting. Collaborated with product owner, tech leads and ops teams to scope requirements and validate design. Identified performance bottlenecks and SLA violations in real-time on microservices.
  3. Implemented hive query support in marketing platform to advanced audience targeting using Hive datasets. Queries were on-demand and near real-time on terabyte scale hybrid text/parquet datasets.
  4. Built reporting on marketing campaign ROI contrasting control and test groups, with APIs delivering dynamically generated Excel sheets from user provided templates using Apache POI. Skills - Full-stack development, Data Engineering, Microservices (Java, Scala, Spring, Angular, Apache Spark, Python) (Promoted to Senior Software Engineer in 2023)
Software Engineering Intern - VISA Summer Intern
Jun 2019 - Aug 2019

Built experimental feature to read decision tree models and execute at scale on audience targeting platform.

  1. Enabled data scientists to run pre-trained models on large datasets in a self-service fashion.
  2. Added data pipeline stage, APIs, UI components. Handled production issues and delivered product demo. Skills - Full-stack development with Scala, Java, TypeScript, Debugging, Software Testing, Large Information Systems
Software Engineering Intern - Autodesk Semester Intern
May 2018 - Oct 2018

Web development and ML model evaluation for Autodesk Design Graph, a ML platform for 3D design data.

  1. Contributed functional and test code to Express APIs in Node.js for the supporting cloud framework.
  2. Evaluated ML algorithms on evaluation datasets and produced analysis for multi-label classification problem. Skills: Node.js, Python, REST APIs, Machine Learning

Education

2023 - 2025
Master of Science in Computer Science
Georgia Institute of Technology
GPA: 4.0 out of 4.0

Specializing in Machine Learning and Computing Systems

  • Distributed Systems
  • Machine Learning with Limited Supervision
2016 - 2020
Bachelor of Computing (Hons)
National University of Singapore
GPA: 4.28 out of 5.0
  • Double specialization in Artificial Intelligence and Computer Security
  • 49th IT Director for the National University of Singapore Photographic Society (NUSPS)
  • Semester abroad at ETH Zürich studying Information Retrieval and Natural Language Processing

Projects

Accelerating Mixture of Experts Inference
Hugging Face PyTorch
Accelerating Mixture of Experts Inference
Research into using novel pre-fetching strategy to accelerate MoE inference. In progress.
Reducing Sycophancy in Large Language Models
Hugging Face PyTorch
Reducing Sycophancy in Large Language Models
Using adversarial frameworks to reduce sycophancy in LLMs.
Distributed key-value store
Java Paxos 2PC Consistency Primary-Backup
Distributed key-value store
Built a distributed key-value store from scratch using Paxos and 2PC using DSLabs Framework. Supported sharding, replication, consensus and transactions. Closed source code as requested by authors (available on request).
Deeno Search
Angular 15 Apache Spark Scala Redis Spring Boot
Deeno Search
Building a search engine from the basics. I've always wanted to do this.
ctrl-f
Nodejs Tesseract OCR Python
ctrl-f
Search a video for phrases (audio search + video search).
3D Reconstruction using Computer Vision
MATLAB Computer Vision Perspective Projection
3D Reconstruction using Computer Vision
Reconstruct the trajectory of a table tennis ball from footage.
vois
Swift iPad App
vois
An iPad app to help budding musicians, seamlessly track their music progress and seek feedback.cp
API Server Template
ExpressJS Nodejs
API Server Template
Create your API server skeleton using this quick and simple API Server Template
Infinity Book
Java Software Engineering Principles
Infinity Book
AddressBook Desktop Application for recruiters.

Get in Touch

I love hearing from you - Email me!