Shree Joshi

Shree Joshi

Computer science student at UT Dallas, building AI and full-stack applications.

B.S. in Computer Science Graduating May 2027 Dallas, Texas

30+
responsive pages shipped in evPanda phase 1
~70%
lower annual vehicle running cost with evPanda
97%
masked-word accuracy fine-tuning BERT in CS50AI

Education

University of Texas at Dallas

Graduating May 2027

Bachelors in Computer Science Minor in Asian Studies Dallas, TX

Coursework

  • Artificial Intelligence (ongoing)
  • Database Systems (ongoing)
  • Probability & Statistics
  • Software Engineering
  • Object-Oriented Design
  • Operating Systems Concepts
  • Automata Theory
  • Programming Language Paradigms
  • Advanced Algorithm & Analysis
  • Systems Programming in UNIX
  • Digital Logic & Computer Design
  • Computer Architecture
  • Data Structures & Algorithms

Projects

AskMahajan

2026

Professional project

B2B SaaS app providing persistent-memory, RAG-enabled conversational AI chatbots on Telegram and WhatsApp, with integrated inventory management and a human-in-the-loop design.

  • Resilient business service with an administration panel, inventory management, bespoke chatbot generation, ticketing and client data isolation, on Cloudflare Workers.

ReactExpressNode.jsR2Neon PostgreSQL

Action AItem

2026

Professional project

Mentored recreation of the live AI application SpeakXYZ, with guidance from its original creator. AI meeting intelligence: transcription, semantic search, AI chat, authentication, and automated meeting minutes, executive summaries and action items.

  • Cloud-native full-stack application with scalable APIs and data management.

ReactNode.jsCloudflare WorkersR2Neon PostgreSQL

evPanda

2025 to present

Professional project

Lead developer of a full-stack EV marketplace, inspired by government initiatives promoting EV adoption and vehicle conversions, reducing annual vehicle operating expenditure by about 70% while lowering carbon emissions.

  • Phase 1: delivered 30+ responsive pages covering vehicle buy/sell, retrofit, spare parts, repair services, advanced filtering, vehicle comparison, theming, and client/vendor role-based access.
  • Phase 2 (present): AI-powered capabilities, including a custom chatbot/voicebot, RPA and AI-enhanced 3D modelling.

ReactViteTypeScriptNode.jsExpressPostgreSQL/PostGISGitLab CI/CDJest/SupertestAWS EC2Nginx

Harvard CS50AI: Intro to AI with Python

2026

Academic project

10+ hands-on projects covering transformer architecture, deep learning, reinforcement learning, NLP/NLU, knowledge-based systems, probabilistic reasoning and search algorithms in Python.

  • Attention: used a pre-trained BERT transformer to predict masked words with 97% accuracy, and visualized attention patterns across layers.
  • Traffic: a convolutional neural network (CNN) that classifies 43 categories of traffic signs from the GTSRB dataset, with image preprocessing, in TensorFlow/Keras.

Prompt Engineering (Edureka)

2025

Academic project

Built ML models for multilingual sentiment analysis, re-purchase prediction, handwritten-digit GAN reconstruction and recognition, and a LangChain RAG pipeline (web scraping, mpnet retriever, BART, Chroma vector store, GPT-2/3.5/4, Claude and LLaMA integration).

TensorFlowPyTorchLangChainTransformerspandasNumPyMatplotlibNLTKscikit-learnBeautiful Soup

What's the Time There

2024

Academic project

A website on AWS that tells the time at a location, using JavaScript, Express, Handlebars and Node.js with the TimeAPI and MapBox services.

Practice projects

Academic project

Multi-lingual semantic transliteration tool, Ren'Py game, auction and inventory system (Python, SQL), parking management system (Python), text manipulation tool (Java), Word-Search AI (Java).

Skills

skills.s
; Skills, written like code. .section languages ; what I write code in, from low-level to web Python C C++ MIPS Java JavaScript TypeScript Prolog Racket SQL HTML/CSS .section ai ; building software that uses large language models LangChain Prompt Engineering Agentic AI RAG LLM Design TensorFlow scikit-learn NLTK .section assistants ; AI coding tools I work with day to day Claude Code Antigravity GitHub Copilot .section web ; websites and the servers behind them Frontend UI/UX React Vite Express Handlebars Node.js .section cloud ; where it all runs Cloud Docker R2 Linux .section toolbox ; everything else NumPy Matplotlib Selenium Git VS Code Vim Ren'Py Microsoft Office .end

Certifications and clubs

Certifications

AWS Certified Cloud Practitioner badge HarvardX CS50 AI badge
  • AWS Certified Cloud Practitioner
  • AWS Certified Solutions Architect, Associate (ongoing)
  • CS50AI (Harvard OCW)
  • Prompt Engineering, Edureka (audited)

Clubs

  • Women in STEM (WinSTEM)
  • AWS Cloud Club
  • Japanese Language Association (JLA)
  • Korean Language Club (KLC)
  • Comet Satellite Laboratory (CSL)

Languages

I speak seven languages: English, Hindi and Gujarati (native), Korean (intermediate), and Japanese, French and Mandarin Chinese (basic).

Each tile shows the language in its own script, its English name, a romanised spelling and the IPA pronunciation. The bar shows how far along it is, from Basic through Intermediate and Fluent to Native; the A1 to C2 tags are the European CEFR scale.

English

English/ˈɪŋɡlɪʃ/

NativeC2

हिन्दी

HindiHindī/ɦɪndiː/

NativeC2

ગુજરાતી

GujaratiGujarātī/ɡudʒəɾɑːtiː/

NativeC2

한국어

KoreanHangugeo/hanɡuɡʌ/

IntermediateB1–B2

日本語

JapaneseNihongo/nihoɴɡo/

Classroom studyA1–A2

français

FrenchFrançais/fʁɑ̃sɛ/

BasicA1

中文

Chinese (Mandarin)Zhōngwén/ʈʂʊŋ˥ wən˧˥/

BasicA1

Contact

Hiring for an entry-level software engineering role, or curious about a project? Send a message.

sjoshi1075@gmail.com

GitHub  ·  LinkedIn