Hi, I'm Vishvam 👋

Machine learning engineer with a love for full-stack development, compilers, and low-level tech.

Vishvam

About

I design and build AI and ML systems, pipelines, and infrastructure at scale. I enjoy exploring library internals, profiling code, and rebuilding things to understand how they really work. I also run experiments on mini PCs and Raspberry Pis, and love trying out new languages, tools, runtimes, and frameworks. Check out my blog, GitHub and resume for projects and experience.

Work Experience

cs_logo.jpeg
Contentstack 3 yrs

Machine Learning Engineer 1

Dec 2024 — Present
Architected and developed the AI Platform backend (RBAC auth, token-based access, granular and configurable rate limits, onboarding, analytics, audit logs) and led a 5-person team to deliver all required features in 5 months. Co-developed the AI Platform UI (ShadCN, React, Vite) and implemented product facing features such as Multimodal Vault, RAG and LLM entries, enabling scalable AI / GenAI operations. Refactored multiple microservices into a unified service, reducing infrastructure costs by 25%

Associate Machine Learning Engineer

Aug 2024 — Dec 2024
Developed and maintained core microservices and AI features of Brand Kit in FastAPI & NestJS earning the Thank You Award for exceptional performance. Led E2E automation using Page Object Model in Playwright, reducing QA effort by 90%. Received the Above and Beyond Award for the same

Associate Machine Learning Engineer Intern

May 2024 — Aug 2024
Designed, developed, and deployed a cost-efficient LLM fine-tuning pipeline on AWS and fine-tuned Mistral-7B, Llama3-8B, Phi-2, and Gemma-7B with 10-12% improved contextual accuracy. Developed a unique way of automated dataset creation, fine-tuning, and benchmarking using simple YAML files. Evaluating models with BLEU, ROUGE, Jaccard, and Perplexity metrics. Received the Thank You Award for AI related efforts and POCs.

Associate Software Engineer Intern

Jan 2023 — May 2024
Integrated Preact and DeepSignal into Visual Builder SDK, reducing bundle size by 24% and earning leadership recognition. Built a framework agnostic component library using Mitosis with a monorepo setup with Lerna accompanied by numerous npm scripts. Ensured a 100% test coverage for all components created via Mitosis

Education

nu_logo.png
NIIT University BTech. Computer Science
Aug 2020 - Aug 2024

CGPA : 8.7 / 10

Coursework :

  • Data Structures
  • Algorithms
  • Compiler Design
  • Computer Networks
  • Operating Systems
iitm_logo.png
IIT Madras BSc. Data Science and Programming
Aug 2020 - Aug 2023

CGPA : 8.0 / 10

Coursework :

  • DSA in Python
  • OOP in Java
  • Machine Learning Theory and Practical
  • Deep Learning Theory and Practical
  • Reinforcement Learning
  • Linux
  • Modern App Development

Skills

Programming Languages

C / C++ Python Typescript / Javascript Go 🌱 Rust 🌱 Java 🌱

Frontend

HTML, CSS (SCSS, SASS, Tailwind, Bootstrap) React Vue Vite Jinja Flutter 🌱

Tools

Bash, Zsh Git, Github VS Code, Zed, Nano Postman, Insomnia Docker Desktop

Backend

FastAPI Flask Express NestJS MongoDB PostgreSQL SQLite Redis ElasticSearch SurrealDB Docker & k8s AWS

Machine Learning / GenAI

Numpy Pandas Scikit-Learn PyTorch Transformers Langchain Unsloth LoRA TRL AWS Bedrock
Selected Projects

What I’ve Been Building

I’ve worked on all kinds of projects : small tools, ML systems, and random experiments. Here are a few of my favorites.

Proxima

Algorithms

A C++17 implementation of the HNSW index for approximate nearest neighbor search, with SIMD acceleration and flexible performance tuning.

C++ Machine Learning Graph Algorithms SIMD
NeuroGC

Machine Learning

Using deep learning models to optimize Python's garbage collection

Python PyTorch Deep Learning Garbage Collection
SpeakUp

Machine Learning

Realtime speech + facial emotion analysis with ML and FastAPI backend.

FastAPI React MongoDB AWS S3 Wav2Vec2 Emotion Detection
Nanoscript

Programming Language / Compilers

A lightweight interpreted language with custom lexer, parser, and closures.

Python Interpreter Programming Language Creation
Contact

Let's get in touch

Feel free to connect with me on LinkedIn or send me an email at s.vishvam1025@gmail.com. I'll get back to you as soon as I can.