Soumya Suryawanshi

Software Engineering Intern Full-Stack Developer AI/ML Specialization

View Projects Contact Resume
— Faster Lookup
— Packet Delivery
— NLP Accuracy
// 02 About

Who I Am

Professional portrait of Soumya Suryawanshi

Computer Science undergraduate specialising in Artificial Intelligence and Machine Learning, focused on software engineering and full-stack development.

I build and deploy cloud-connected applications — from Flutter mobile apps backed by six Firebase services, to NS-3 network simulations and NLP classifiers. I've worked within a Fortune 500 Agile environment at Reliance Industries Limited and bring hands-on experience across Flutter, Dart, Firebase, Node.js, REST APIs, SQL, C++, and Python.

SIT

B.Tech — Computer Science

AI/ML Specialization

Symbiosis Institute of Technology, Nagpur

Nagpur, India
B.Tech CS '28
Full-Stack · Mobile · ML
// 03 Experience

Where I've Worked

Navi Mumbai Completed

Software Engineering Intern

  • Engineered enterprise mobility solutions for Reliance's Petchem IT division within a Fortune 500, Agile SDLC environment.
  • Built EmpFlow, a cross-platform employee database and workforce management system using Flutter, Dart, Firebase, and Cloud Firestore.
  • Implemented employee CRUD, Firebase Authentication, admin/staff role-based access control, real-time multi-field search, QR generation/scanning, PDF/Excel exports, and FCM push notifications.
  • Designed layered architecture with Provider state management and Firestore real-time listeners; integrated six Firebase services.
  • Optimised Firestore collections, indexed queries, and stream caching; internal testing showed ~85% faster lookup, 150+ records validated, and <1 sec QR scan-to-profile time.
  • Used Git/GitHub for version control and deployed the web application through Firebase Hosting.

EmpFlow is not an officially endorsed Reliance Industries product. This internship is completed; I am not currently employed at Reliance.

Completed

Machine Learning Intern

  • Architected an end-to-end face recognition pipeline using PCA, ANN, Python, NumPy, and OpenCV.
  • Improved robustness across varying lighting and poses through preprocessing and hyperparameter tuning while reducing inference latency.
// 04 Selected Projects

What I've Built

Signature Case Study · 2026

EmpFlow

Cloud-Based Employee Database & Workforce Management System

Personal project built during internship. Not an officially endorsed Reliance Industries product.

Flutter Dart Firebase Cloud Firestore Provider FCM

A cross-platform workforce management system — employee CRUD, role-based access control, real-time multi-field search, QR identity workflow, PDF/Excel export, and FCM push notifications. Layered architecture using Provider state management with six Firebase services integrated.

85% Faster Lookup
150+ Records Validated
<1s QR Scan-to-Profile

Results from internal testing — not independent benchmarks

Interactive Architecture — hover / click / tap a node to learn its role

Screenshots / screen recording — add your own media here

Replace public/images/empflow-screenshots/
Research Project · 2025

Hybrid MANET Routing Protocol Simulation

NS-3 C++ Linux

Simulated AODV and DSDV routing protocols across 50+ node mobile ad-hoc networks, achieving ~94% packet delivery ratio. Analysed latency, throughput, and routing overhead; work contributed to a published patent application.

94% Packet Delivery Ratio
50+ Nodes Simulated

Simulation results — no claim that either AODV or DSDV is superior in all conditions

⚠ Illustrative animation — not NS-3 simulation output

ML Project · 2025

Fake News Detection System

Python NLP TF-IDF Logistic Regression Scikit-learn

Built an NLP classifier achieving ~91% accuracy across 8,000+ articles using TF-IDF vectorisation, tokenisation, and stopword removal. Validated performance using cross-validation and precision-recall analysis.

91% Accuracy
8,000+ Articles Evaluated

Results on held-out test set with cross-validation

⚠ Illustrative diagram — not a live model inference

ML Pipeline · iStudio · 2025

Face Recognition Pipeline

Python NumPy OpenCV PCA ANN

Architected an end-to-end face recognition pipeline using PCA for dimensionality reduction and an ANN classifier. Improved robustness across varying lighting and poses through preprocessing and hyperparameter tuning while reducing inference latency.

⚠ Illustrative technical diagram — no accuracy figures are claimed

Published Patent Application

Computer-Implemented MANET Routing Performance System

Application No. 202521127182 Published

This is a published patent application, not a granted patent.

// 05 Technical Skills

My Toolkit

Hover a technology to see which projects use it.

Languages

Software Engineering

Mobile & Backend

Web & Databases

Systems & AI

// 06 Certifications

Credentials

Google · Coursera

Google AI Professional Certificate

Deloitte · Forage

Deloitte Technology Virtual Job Simulation

Red Hat Academy

Red Hat System Administration I

RH124
Red Hat Academy

Python Programming

Red Hat Academy
// 07 Contact

Let's Connect

Reach out directly

soumyasuryawanshi139@gmail.com
+91 7709299542 linkedin.com/in/soumyasuryawanshi github.com/soumyasurya1309