Home

About

Journey

Projects

Skills

Contact

Home About Journey Projects Skills Contact

Hi there! 👋

I'm Abhishek Jani.

Based in New Brunswick, NJ, I'm a Software Engineer & MS CS student at Rutgers. I love building distributed systems, ML pipelines, and cloud-native applications. Passionate about applied AI and shipping reliable production code.

Get in Touch!

let's build something ✨

ML / AI Engineer

Node.js

React

Python

JavaScript

PyTorch

AWS

Docker

ABOUT

I'm a passionate software engineer and MS Computer Science student at Rutgers University (GPA: 3.78), currently based in New Brunswick, NJ.

My work spans full-stack development, cloud infrastructure, and machine learning — from shipping production backends that handle 2K+ requests/minute to fine-tuning BioBERT models achieving 95.6% F1-scores on clinical NLP tasks.

I thrive at the intersection of systems thinking and applied AI — building fault-tolerant distributed systems, designing serverless self-healing pipelines on AWS, and creating observability platforms powered by LLMs. I care deeply about code that is reliable, scalable, and actually ships.

2+ Years Experience (incl. Internships)

5+ Major Projects

3.78 MS GPA @ Rutgers

20+ Technologies

My Journey

Journey Timeline

Software Engineering Intern @ Attention Group LLC

Jun 2025 – Dec 2025

Shipped production Node.js/Express + MongoDB backend with 30+ REST endpoints serving 2K+ req/min at 99.9% uptime. Built image-generation workflows with React + OpenCV. Implemented OAuth 2.0 for Twitter API integration, reducing login errors by 85%.

Remote / New Jersey, USA

MS, Computer Science @ Rutgers University

Sep 2024 – May 2026

GPA: 3.78/4.0. Coursework: Machine Learning, NLP, Computer Vision, Deep Learning, Database Systems, Interpretable AI, Data Structures & Algorithms.

New Brunswick, NJ, USA

Technical Support Engineer L1 @ Sophos

Aug 2023 – May 2024

Resolved 200+ technical incidents for IPSec, SSL VPN, VoIP modules. Reduced resolution time by 18%, first-response time by 15%. 95% customer satisfaction rating.

India

Research Intern @ Ganpat University

Jan 2023 – May 2023

Improved image classification accuracy by 25% on EMOTIC dataset using hybrid CNN (AlexNet + VGG19). Achieved 73.6% validation accuracy on emotion recognition task.

Gujarat, India

BTech, Computer Engineering @ Ganpat University

Jul 2019 – May 2023

GPA: 8.97/10. Built a strong foundation in algorithms, data structures, computer networks, and software engineering.

Gujarat, India

PROJECTS

Things I've built that I'm proud of

FastAPI · React Flow · Gemini AI

Sentinel

AI-Powered Network Health Monitor

Customer-centric observability platform using FastAPI + async SSE processing real-time event streams. Google Gemini AI generates root cause analysis before customer impact. Z-score anomaly detection (2.0σ threshold) with 30-point rolling baselines across O(customers × services) matrix updated every 15s.

FastAPI React Flow Gemini AI SSE Anomaly Detection

Python · aiohttp · Flask

Adaptive Reverse-Proxy

High-Performance HTTP Load Balancer

HTTP reverse proxy with three adaptive routing algorithms (Round-Robin, SMA, EWMA) reducing tail latency by 35%. Persistent connection pooling with async I/O — 60% latency reduction, 45% throughput improvement. Real-time Plotly.js dashboard with p95 tail latency trends.

Python aiohttp Docker Plotly.js Load Balancing

AWS Lambda · Step Functions · CDK

Runbook Ranger

AWS Auto-Remediation Bot

Serverless self-healing system using AWS Lambda, Step Functions, and EventBridge to detect and resolve incidents automatically — MTTR from minutes to seconds. Production-grade safety: distributed resource locking, idempotency keys, blast-radius limits. Runbooks-as-Code engine in Python/YAML via AWS CDK.

AWS Lambda Step Functions EventBridge CDK IAM

Python · PSI · Data Contracts

Data Reliability Guardrails

Fault-Tolerant Pipeline Validation

Fault-tolerant validation system enforcing strict data contracts (Schema, Freshness, Distribution) preventing silent data failures in batch pipelines. PSI algorithm + volume checks detect distribution shifts with <5ms latency. 98% automated detection rate via custom benchmarking with failure injection.

Python PSI Schema Validation Data Contracts

BioBERT · Hugging Face · NLP

ADE Detection with Explainable Counterfactuals

Clinical NLP + BioBERT

Fine-tuned BioBERT on ADE-Corpus V2 (11,914 clinical sentences) achieving 95.6% F1-score (95.3% precision, 96.2% recall). Counterfactual analysis revealed 76.5% prediction flip rate in ADE-positive cases vs 1.9% in non-ADE. Evaluated lexical robustness via RxNorm API drug name substitution across 8,448 sentences.

BioBERT Hugging Face PyTorch NLP Counterfactuals RxNorm API

SKILLS

Languages

Python

C++

Java

JavaScript

TypeScript

SQL

Frameworks

React

Node.js

Next.js

FastAPI

Flask

Django

Express

ML / AI

PyTorch

TensorFlow

Scikit-Learn

OpenCV

BioBERT

Pandas

NumPy

Cloud & DevOps

AWS Lambda

Step Functions

Docker

EventBridge

CDK

Prometheus

Grafana

Databases

PostgreSQL

MongoDB

MySQL

ChromaDB

Redis

Tools & More

Git

Linux

REST API

OAuth 2.0

Microservices

CI/CD

Distributed Sys.

EDUCATION

🎓

Master of Science, Computer Science

Rutgers University

GPA: 3.78 / 4.0

Sep 2024 – May 2026 · New Brunswick, NJ

🏛️

Bachelor of Technology, Computer Engineering

Ganpat University

GPA: 8.97 / 10

Jul 2019 – May 2023 · Gujarat, India

HONORS

🏆

Amazon ML Summer School

Amazon · Jul 2022

Selected participant for intensive ML program covering Supervised Learning, Deep Neural Networks, Probabilistic Graphical Models, Reinforcement Learning, and Causal Inference.

🛡️

Sophos Certified Engineer

Sophos · Sep 2023

Certified in Sophos Firewall v19.5, v20.0 and MDR — demonstrating expertise in network security protocols.

GET IN TOUCH

Let's build something amazing together

Email abhishekjani075@gmail.com LinkedIn Connect with me GitHub See my code