Skip to content
View Preksha2's full-sized avatar

Highlights

  • Pro

Block or report Preksha2

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Preksha2/README.md

Hey, I'm Preksha

MS in Computer Science @ Northeastern University (Khoury College) | Boston, MA

Currently a Teaching Assistant for CS 3650 (Computer Systems), helping 400+ students navigate memory management, concurrency, and low-level systems programming.

What I work with

Languages: Python, Java, Go, C++, C, SQL, CUDA, Shell/Bash

ML/AI: PyTorch, FAISS, RAG Pipelines, LLM Integration, NLP, CUDA Kernel Optimization

Backend & Infra: FastAPI, Flask, Node.js, Docker, Kubernetes, AWS (EC2, S3), Redis, CI/CD

Data: PostgreSQL, MySQL, MongoDB, FAISS

What I've been building

  • Real-Time AI Document Engine — Production RAG pipeline over 50K+ event log chunks using FAISS + LLM, with four-dimensional evaluation (relevance, groundedness, reliability, safety) and horizontal scaling via Docker + Nginx

  • Multi-Object Tracking System — Real-time tracking pipeline using YOLOv8 with custom CUDA kernels for NMS and bounding box regression, achieving 35 FPS on 1080p streams

Previously

  • ML Intern @ ByteCitadel — Distributed ML pipeline on Kubernetes processing 100M+ records/day, custom CUDA kernels for 40% throughput improvement, P95 latency reduced from 500ms to 190ms

  • AI/CV Researcher @ Ahmedabad University — Traffic monitoring with multi-object tracking, evaluated 5+ SOTA models on 50K+ annotated frames

  • SDE Intern @ Warble Solutions — Microservices with Redis caching, CI/CD automation, zero-downtime Kubernetes deployments

Let's connect

LinkedIn Email

Pinned Loading

  1. realtime-ai-document-engine realtime-ai-document-engine Public

    Production-style RAG pipeline for real-time document querying using FAISS, FastAPI, and LLMs

    Python

  2. research-paper-etl-pipeline research-paper-etl-pipeline Public

    Full-stack AI-powered ETL pipeline using LangGraph agents to automatically traverse citation chains, extract metadata via LLM, and build structured research datasets with Pinecone and FastAPI

    Python