Akanksha Garg

Welcome to My Portfolio

I'm Akanksha Garg

Data Engineer | AI Enthusiast | Software Engineer

πŸ’Ό

7 Years in Business

Operations & Analytics

↓
πŸš€

The Evolution

Didn't just pivot - I evolved

↓
πŸŽ“

MS @ RIT

Teaching myself CS fundamentals while pursuing Master's

πŸ€–

AI & Data Engineering

Building systems that matter

πŸ’‘

Relentless Learning

Every algorithm from scratch

↓
🌟

My Mission

Shape the world being redefined by technology

Akanksha Garg

MS Computer Science @ RIT | Specialization: AI and Big Data

About Me

Akanksha Garg

I am a software engineer and data systems builder with experience spanning industrial analytics, machine learning, and scalable backend development. My work focuses on designing systems that transform real-world data into actionable intelligence, combining engineering rigor with applied problem solving.

Before pursuing my master's in Computer Science at Rochester Institute of Technology, I spent several years leading analytics and optimization initiatives in a large-scale industrial environment, where I applied statistical modeling, process analysis, and data engineering to improve production efficiency, yield performance, and operational reliability. That experience shaped my interest in building intelligent systems that operate under real-world constraints.

Currently, I am developing PantryPal, an AI-powered smart kitchen platform that integrates backend systems, data modeling, and computer vision to automate inventory tracking and personalized meal recommendations. My technical interests include machine learning systems, data engineering, distributed architectures, and applied AI in production environments.

I enjoy working on problems that sit at the intersection of software engineering and intelligent decision-making, particularly where system design, data pipelines, and optimization come together.

Technical areas I work in:

  • β€’ Backend and data systems development (Python, SQL, APIs, PostgreSQL)
  • β€’ Machine learning and analytics (feature engineering, clustering, predictive modeling)
  • β€’ Cloud and distributed infrastructure (AWS, scalable architectures)
  • β€’ Applied optimization and real-world data modeling

I am actively seeking software engineering and machine learning internship opportunities where I can contribute to building scalable, intelligent systems.

Rochester, NY 14623

(585) 506-1121

akankshagarg0063@gmail.com

Technical Expertise

Programming

Python SQL JavaScript C++ MATLAB Java HTML/CSS

Machine Learning

PCA DBSCAN Transformers (BERT) Self-Supervised Learning (DINOv2) Feature Engineering Model Evaluation Decision Trees Random Forests Genetic Algorithms

Data Engineering

ETL Pipelines Time-Series Analysis Data Modeling REST APIs Statistical Modeling NumPy Pandas Scikit-learn

Cloud & Databases

PostgreSQL Supabase AWS (EC2, S3, IAM) System Design

Tools

Git GitHub Docker Linux React Next.js PLC/SCADA

Professional Experience

Founder & Software Engineer Β· Self-employed

PantryPal – AI Smart Kitchen Platform Β· Rochester, New York

Feb 2025 - Present Β· 1 yr 1 mo

Designing an AI-powered pantry management platform converting inventory data into personalized recipe and grocery recommendations.

AI Inventory Recognition Pipeline (Python, Computer Vision, Deep Learning – Planned)

  • Designing computer vision–based ingredient detection system for automated pantry recognition using image capture.
  • Developing end-to-end pipeline: image acquisition β†’ object detection β†’ inventory update β†’ recommendation engine.

Recommendation & Meal Planning Engine (Python, SQL, Backend APIs)

  • Implementing rule-based and data-driven recommendation logic to generate recipes from available ingredients.
  • Designing scalable database schemas for inventory tracking and user personalization.

Platform & Infrastructure

  • Developing backend services using Supabase (PostgreSQL), authentication workflows, and REST API integration.
  • Building frontend interfaces using modern web technologies with cloud deployment architecture.

Technologies:

Python Computer Vision Deep Learning SQL Supabase REST APIs Next.js TypeScript

Garg Fertilizers Β· Full-time

Rice Milling Division (M/s Radha Krishna Foods) Β· Faridkot, Punjab, India

Jun 2017 - Jul 2024 Β· 7 yrs 2 mos

Led data-driven optimization initiatives across industrial production operations, applying analytics, process modeling, and operational engineering to improve yield, efficiency, and cost performance.

Optical Sorting Optimization (Sortex Systems)

  • Analyzed defect classification and batch-quality data using Python and statistical modeling to identify factors affecting yield and product grading.
  • Optimized sorting sensitivity thresholds and moisture conditions, improving head rice recovery by 8–12% while reducing rejection losses by ~10%.
  • Established data-driven calibration procedures for consistent premium-grade output aligned with industry quality standards.

Industrial IoT Analytics for Milling Operations

  • Processed PLC/SCADA runtime and downtime logs using time-series analysis to identify bottlenecks and inefficiencies in production workflows.
  • Developed utilization and energy efficiency metrics, increasing machine utilization by 12–18% and reducing unplanned downtime by ~15%.
  • Supported preventive maintenance planning through operational data insights.

Fluidized Bed Cooling Optimization

  • Modeled relationships between grain temperature, moisture variability, and breakage rates using MATLAB and statistical analysis.
  • Identified thermal instability as a major contributor to post-milling breakage and supported deployment of controlled cooling systems.
  • Reduced breakage by ~7–10% and improved downstream sorting efficiency.

Procurement & Cost Modeling

  • Built time-series pricing models using Python and SQL to optimize procurement timing and supplier selection.
  • Developed cost-per-ton production models integrating energy consumption, yield, and input pricing variables.
  • Reduced procurement costs by ~5–9% and improved margin predictability.

Technologies:

Python SQL MATLAB Statistical Modeling Time-Series Analysis PLC/SCADA Industrial Automation

Research Intern

INMAS, Defence Research and Development Organization (DRDO) Β· Delhi, India

Jan 2017 - May 2017 Β· 5 mos

Conducted cognitive workload and brain-signal analysis experiments using EEG data acquisition and processing frameworks.

EEG Signal Processing and Cognitive Workload Analysis (MATLAB, EEGLAB, OpenViBE, NASA-TLX)

  • Processed EEG signals using EEGLAB for artifact removal and feature extraction.
  • Applied NASA-TLX methodology to evaluate cognitive workload across experimental tasks.
  • Developed experimental data analysis pipelines for neuroscience research evaluation.

Virtual Environment Simulation for Human Performance Studies (Unity3D, Java)

  • Built virtual simulation environments to study human interaction and task performance.
  • Integrated experimental scenarios with physiological data collection workflows.
DRDO Internship Certificate

Industrial Engineering Intern

Philips - Mohali, India

2015

  • Manufacturing cost reduction of Compact Fluorescent Lamps; achieved a 7% cost reduction in choke coil production.
Philips Internship Certificate

Co-Founder - Har Hath Kalam (NGO)

Community Outreach & Program Management

2015 - 2018

  • Managed finances and coordinated cross-functional teams to design and execute community outreach programs.

Featured Projects

Data Engineering
663K Records
EmergencyIQ

Production ETL System

EmergencyIQ Platform

45s β†’ <1s query time
73% memory reduction

Built complete ETL pipeline processing 663K emergency dispatch records. Implemented DBSCAN clustering algorithm from scratch (no sklearn) with custom batch processor achieving 73% memory reduction. Optimized queries from 45s to under 1s, resulting in 25% staffing efficiency improvement.

Python SQLite DBSCAN PCA Pandas
Interactive
Practice Tool Open Source
Interview Prep Simulator

Interactive mock interviews

Interview Prep Simulator

Timed practice sessions
Mock interviews & feedback

Interactive simulator to practice coding and behavioral interviews with timeboxed challenges, feedback and scoring. Open-source - ideal for sharpening interview skills under pressure.

JavaScript React Node.js WebSockets
Machine Learning
76K Images
Plant Disease

Deep Learning Model - Under Construction

Plant Disease Detection

95% accuracy
88 disease classes
DINOv2 transformer

Developed deep learning model using DINOv2 transformer achieving 95% accuracy across 88 plant disease classes on 76K images. Solved overfitting through 2-stage training strategy (frozen backbone β†’ fine-tuning) with custom augmentation pipeline. Deployed interactive web interface with Gradio.

PyTorch DINOv2 Transfer Learning
Data Engineering
600K Coordinates 99.86% Accuracy
GPS

GPS Processing System

GPS Analytics Pipeline

99.86% accuracy
Custom NMEA parser

Engineered custom NMEA parser from scratch after existing library (pynmea) failed, achieving 99.86% accuracy (1 error in 10,780 lines) on 600K coordinates. Optimized processing for 79% speed improvement. Automated KML generation reducing manual work from 8 hours to instant processing.

Python Custom Parser Regex KML
Machine Learning
1,197 Reports 72% Accuracy
Weather

Custom ML Algorithms

Weather Prediction System

Decision trees from scratch
Random forests custom
342 model configurations

Built decision tree and random forest algorithms from scratch (no sklearn) achieving 72% accuracy on temperature/precipitation prediction. Scraped 1,197 CF6 weather reports with 100% success rate. Engineered 34 features including circular wind direction transforms. Tested 342 model configurations for optimization.

Custom ML BeautifulSoup Feature Engineering Web Scraping
Research
Comparative Analysis Cross-Domain
Evolutionary Design

Cross-Domain Analysis

Advances in Evolutionary Design

NASA ST5 Antenna Evolution
Xenobots Bio-Design
Fitness Function Comparison

Comparative research analysis examining how Genetic Algorithms solve complex engineering problems across vastly different domains: aerospace (NASA's evolved X-band antennas for ST5 mission) and bioengineering (living reconfigurable organisms). Investigated representation strategies, fitness evaluation methods, and robustness mechanisms. Explored how the same algorithmic framework adapts to domain-specific constraints, from electromagnetic optimization to biological realizability, highlighting GA versatility in high-dimensional, non-linear design spaces.

Genetic Algorithms Multi-objective Optimization Aerospace Engineering Bioengineering Evolutionary Computation
Renewable Energy
Team Project Multi-Module System
Road Power Generation System

Self-Sustained Energy System

Road Power Generation System

Wind: 60-90W output
Kinetic: 2.3kW daily
Piezo: 240kW/km potential
Automated street lighting

Designed multi-module renewable energy system harvesting highway energy through wind turbines (VAWT), kinetic rollers, and piezoelectric sensors. Implemented 8051 microcontroller-based automation with IR/LDR sensors for intelligent street lighting. System demonstrated complete energy self-sufficiency: generation β†’ storage β†’ consumption cycle with zero external power requirements. Undergraduate research project at Thapar University exploring distributed renewable infrastructure.

Renewable Energy VAWT Design Piezoelectric 8051 MCU IoT Sensors Power Electronics

Har Hath Kalam

Co-founded an education-focused NGO supporting out-of-school children through bridge learning, mentorship, and community engagement, helping learners transition into formal education and build confidence for long-term growth.

You can become a changemaker by making a fundraiser for this organization. Click here

Education & Certifications

Master of Science in Computer Science

Rochester Institute of Technology

College Transcript

Expected Dec 2026 | Graduate GPA: 3.6 / 4.0

Professional Certificate - Web Development (Ongoing)

IITM Pravartak | MERN Stack | Java

2025 - Present

React, Node.js, MongoDB, Express.js, Java

Technical Certifications

Bachelor of Engineering

Electronics Instrumentation & Control

2017

Thapar Institute of Engineering & Technology, Punjab, India

College Transcript

National Talent Search Scholar (NTSS)

National Level Scholarship by NCERT, India

National Council of Educational Research and Training (NCERT)

Awarded 2009 | Top 0.16 % Nationwide

NTSS Certificate

🌸 Grace Hopper Celebration 2025

Stories That Stay With Me

Stepping into the Grace Hopper Celebration felt like entering a space shaped by hope, courage, and collective possibility. Thousands of women gathered not just to talk tech, but to share stories of growth, resilience, and community.

GHC Expo Floor
The vibrant expo floor at McCormick Place
Networking at GHC
Connecting with inspiring women in tech
GHC Experience
Celebrating innovation and community

Keynote Speakers

  • Jeanne Sparrow - Navigating change with honesty and grace
  • Brenda Darden Wilkerson - Inclusion as a shared responsibility
  • Vivian Tu - STRIIP method for financial confidence
  • Erin Coupe - Finding clarity in small daily moments
  • Adam Cheyer - AI evolution from systems that know to systems that do

Meaningful Connections

Nicole Kaiser - Grounded resume advice
Joveria Asif (Morgan Stanley) - Recognized leadership traits I often overlook
Nafisa Ali Amir (Audible) - Reframed career pivots as intentional growth
Kelcie Dinkel (Cloudflare) - Encouraged me to aim higher with AI work
Vedika Baid (Goldman Sachs) - Humility and genuine curiosity that left a mark
Sripriya Simhadri (InterSystems) - Appreciated my self-learning journey
Sai Iyer (Prudential) - Thoughtful connection over background
Jayasree K (Amazon) - Reminded me reinvention is always possible
Aishwarya Teegulla (Bloomberg) - Pushed me to reflect on my "why"

Special Moments

  • Meta's Women Shaping Wearables - Hind Hobeika shared her remarkable founder journey and took time to hear my story. We even took a photo I'll always cherish.
  • Kelly Ingham - Expanded our imagination of what's possible in wearables
  • Closing Celebration - Voices like Angelica Ross, Kellie Gerardi, and Betsy Tong brought messages of courage, authenticity, and collective progress

"As I left McCormick Place, I carried with me not just business cards or LinkedIn connections - but stories, lessons, and hope."

GHC 2025 reminded me that no journey is linear, no background limits your potential, and no dream is too unconventional when women choose to lift one another.

I'm beyond grateful to everyone I met for their time, kindness, and wisdom. πŸ’œ

Let's Build Something Extraordinary

Whether you're looking for a passionate data engineer, an AI enthusiast ready to tackle complex challenges, or someone who brings 7 years of business acumen to tech - I'd love to connect.

Open to Opportunities

Full-time roles, internships, and contract work in Data Engineering, Machine Learning, and AI

  • πŸ’Ό Data Engineering & ETL Pipelines
  • πŸ€– Machine Learning & AI Solutions
  • ☁️ Cloud Architecture (AWS, GCP)
  • πŸ“Š Big Data & Analytics

Let's Collaborate

Interested in discussing innovative projects, research ideas, or just connecting over tech?