Skip to content

Maharshi Jani

@maharshijani05

I’m an AI enthusiast passionate about building intelligent systems that solve real-world problems. Skilled in ML, Generative AI, LangChain, NLP and Computer Vision.

I’m an AI enthusiast passionate about building intelligent systems that solve real-world problems. Skilled in ML, Generative AI, LangChain, NLP and Computer Vision.

Skill iconPython
Machine Learning
Deep Learning
Computer Vision
Natural Language Processing

Software Engineer Intern, Softmaxai

Nadiad, India

👋 Hi, I'm Maharshi Jani

🚀 Passionate about building impactful AI-powered solutions that bridge technology and real-world needs.
🎯 I specialize in Generative AI, ML Engineering, and AI-Augmented Applications.
🌐 Exploring the intersections of machine learning, civic tech, and finance automation to build systems that think, adapt, and empower users.


🧠 Featured Projects

🏛️ CivicMind – AI-Powered Civic Policy Simulator

Solo Hackathon Project | Awarded for Originality & Uniqueness (HolboxAI Hackathon)
Tech Stack: LLMs (Groq/OpenAI), LangChain, Multi-Agent AI, Streamlit

CivicMind simulates real-world governance using AI agents that debate, evaluate, and score public policies. Users can upload a policy, and the system triggers a roundtable-style simulation with agents representing different citizen archetypes.

  • 🧠 LLM-powered agents with distinct personas, memories, and reasoning abilities.
  • 🗳️ Dynamic tone analysis, JudgeBot scoring, and interactive UserAgent participation.
  • 📊 Structured impact assessment with feasibility, sentiment, and scoring breakdowns.

🔗 Live Demo | GitHub


🤖 MLPilot – Your Personal ML Co-Pilot

Team Project | ML Automation Tool for Developers and Students
Tech Stack: Streamlit, Scikit-learn, Pandas, LLMs

MLPilot simplifies machine learning workflows by auto-generating ready-to-run pipelines for classification and regression using just a CSV.

  • ⚙️ Automatic data cleaning, model selection, training, and evaluation.
  • 📋 One-click code export for reproducibility and deployment.
  • 💬 LLM-powered explanations for model choices and results.

🔗 Live Demo | GitHub


💰 BudgetPro – Smart Financial Planning Assistant

Team Project | AI-Powered Personal Finance App
Tech Stack: XGBoost, LangChain, MongoDB, Python, Groq

BudgetPro is a full-stack AI assistant that helps users track, analyze, and predict their expenses and budgets using machine learning and natural language understanding.

  • 💡 XGBoost-based transaction categorizer with over 92% accuracy.
  • 📈 User clustering and personalized budget prediction with savings recommendations.
  • 🧠 Conversational assistant built with LangChain + Groq for real-time financial insights.

🔗 GitHub


🧑‍💻 What I’m Good At

  • 💡 AI/ML: XGBoost, Clustering, LSTMs, Transformers, RL, Model Deployment
  • 🧠 Generative AI: LangChain, Prompt Engineering, OpenAI, Groq
  • 🛠️ Tech Stack: Python, Streamlit, Flask, Node.js, MongoDB, Docker, AWS
  • 🌍 Domains: Civic Tech, EdTech, FinTech, Habit Engineering

📫 Let’s Connect


“I don’t just build models — I build systems that reason, adapt, and deliver value.”