01 /About

Engineering Intelligence,
One Model at a Time

I'm a Computer Science undergraduate at Guru Jambheshwar University who became obsessed with one question: what happens when you give data a voice?

From building Random Forest regressors that forecast market movements to co-developing AI triage agents for clinical workflows, I gravitate toward problems where machine learning creates measurable, real-world impact — not just benchmark scores.

Outside of model training loops, I lead teams in national hackathons, represent India at international student conferences, and build hardware automation solutions for social impact.

joyashatri@gmail.comB.Tech CSE · 2024–28
5+
ML Projects Built
3
Hackathons / Competitions
100+
Teams Competed Against
25%
Consultation Load Reduction (AI Triage)
joyash@dev ~
$ python train.py --model rf --data markets
→ Training complete. RMSE: 0.043
$ python predict.py --next-day
02 /Featured Work

Projects That Ship

Every project is a production-minded experiment — built to measure impact, not just check boxes.

01

Stock Price Predictor

Market Intelligence via ML

Engineered a production-grade Random Forest regression pipeline that forecasts next-day closing prices with sub-5% error. Applied time-series feature engineering — MA-10 moving averages, lag variables, volatility metrics — on 5+ years of historical data.

~4.3% avg RMSE on test set
  • Random Forest Regressor
  • Time-series forecasting
  • Feature engineering
  • Matplotlib visualizations
PythonScikit-learnPandasNumPyMatplotlib
02

Abhilasha AI Medical Agent

Agentic AI Hackathon at Ulster University, UK

Co-built an NLP-driven medical assistant that pre-screens patient queries and routes clinical cases intelligently. The system demonstrated a 25% projected reduction in basic consultation load in prototype testing, selected for presentation at Ulster University.

~25% reduction in consultation load
  • AI-powered patient intake
  • Query classification
  • Clinical workflow automation
  • Ulster Uni collaboration
PythonNLPAI AgentsHealthcare AI

AI Research Agent

03

Personal Research Assistant for Academic Papers

Developed an AI-powered research assistant in Jupyter Notebook that automates data cleaning tasks such as handling missing values, duplicate removal, outlier detection, feature standardization, and dataset preparation for analysis

PythonPandasNumpyMatplotlib

Sanitary Pad Vending Machine

04

Arduino Automation for Social Impact

Designed and prototyped an Arduino-powered automated vending system targeting menstrual hygiene accessibility in schools and public institutions. Developed a scalable deployment concept built for low-resource environments.

ArduinoHardwareAutomationIoT
03 /Skills

Tech Stack & Toolkit

{ }Languages
PythonC++JavaScriptSQLHTML/CSS
ML / AI
Scikit-learnPandasNumPyMatplotlibRandom ForestTime-series ForecastingNLPFeature Engineering
Tools & Frameworks
PyPDF2pyttsx3TkinterArduinoGitGitHub
Concepts
Supervised LearningData PreprocessingModel EvaluationIoT AutomationAgile / HackathonsTechnical Communication

Proficiency

Self-assessed based on project experience and depth of use.

Python / ML85%
Data Analysis80%
Model Building78%
Hardware / Arduino65%
Team Leadership90%

Currently Exploring

Deep LearningPyTorchLLM Fine-tuningFastAPIDocker
04 /Leadership & Experience

Beyond the Code

Team Lead

Smart India Hackathon

2024Competition

Digital Platform for Cultural Heritage Site

  • Directed a 5-member cross-functional team to design and ship a scalable digital heritage engagement platform end-to-end
  • Architected the system with a modular UI and integrated data management layer for real-time visitor interactions
  • Outperformed 100+ competing teams to advance to the internal presentation round, pitching to an evaluation panel
Selected · Internal Round

AI Healthcare Prototype Developer

Ulster University, UK

2024International Collaboration

"Abhilasha" AI Medical Agent

  • Co-engineered an AI triage assistant to intelligently classify and route inbound patient queries at a preliminary stage
  • Prototype demonstrated a projected 25% reduction in basic consultation load in simulated clinical workflows
  • Shortlisted for the presentation round based on solution feasibility, technical depth, and impact potential
Shortlisted · Presentation Round

Harvard Project Delegate

VCONF '26 — Asian & International Relations

2026International Conference

Asia-Pacific Policy & Technology Track

  • Selected as an Indian delegate to represent perspectives on technology policy at the Harvard VCONF '26
  • Engaged with international delegations on emerging technology, AI governance, and cross-border digital cooperation
Selected Delegate
Education

B.Tech — Computer Science & Engineering

Guru Jambheshwar University of Science & Technology, Hisar

2024 – 2028

Currently enrolled

05 /Contact

Let's Build
Something Smart

I'm actively seeking internships, research collaborations, and engineering roles where machine learning meets real impact. If you're working on interesting AI/ML problems, I'd love to connect.

Open to internships, research roles, and project collaborations