Satyam Sharma
LinkedIn | GitHub | Portfolio
Education
University of Mumbai
Mumbai, India
Bachelor of Science in Computer Science
Jun 2021 – May 2024
– Relevant Coursework: Object-Oriented Programming, Computer Networks, Operating Systems, Artificial
Intelligence, Python and Java Programming, Data Structures and Algorithms.
Experience
ML Intern
Dec 2024 – May 2025
Kredit (Startup)
Remote
– Designed an AI-driven credit scoring system for gig economy workers, improving loan accessibility through
automated, data-driven risk assessment.
– Improved the core credit model’s predictive accuracy by 40% (R2 ) by leveraging feature engineering and data
preprocessing pipelines with TensorFlow/Keras.
Projects
Nexa – DeFi Yield Aggregator
View on GitHub
– Created a full-stack decentralized application using a Next.js and TypeScript frontend and a
Node.js/Express.js backend for multi-chain yield aggregation on the NEAR protocol.
– Implemented a responsive UI with Tailwind CSS and shadcn/ui, featuring an advanced, sortable data table built
with TanStack Table for data visualization.
– Integrated external financial APIs (DefiLlama, 1Click Intents) to fetch, rank, and generate cross-chain
transaction plans in real-time.
– Constructed the complete end-to-end transaction lifecycle by integrating the NEAR Wallet Selector for on-chain
signing and implementing a multi-state modal with real-time status polling.
CogniCart – AI E-commerce Analytics Platform
View on GitHub
– Built a full-stack analytics dashboard using FastAPI for a high-performance, asynchronous REST API and
Next.js for a responsive, server-side rendered frontend.
– Architected interactive CRUD interfaces with TanStack Table, featuring client-side state management for UI
updates, dynamic sorting, and filtering.
– Formulated a robust backend with SQLAlchemy and Pydantic for data integrity, and devised an AI sales agent
with a RAG pipeline for natural language querying.
– Containerized the application using Docker and Docker Compose to simplify the deployment workflow.
CryptexAI – Cryptocurrency Price Prediction Platform
View on GitHub
– Architected a predictive analytics platform using deep learning models (LSTM, GRU, Transformer) to forecast
cryptocurrency prices from historical time-series data.
– Implemented a full-stack system with a FastAPI backend serving ML inferences and a React frontend visualizing
real-time market data via WebSockets.
– Devised a custom backtesting engine to validate trading strategies and leveraged Optuna for automated
hyperparameter tuning, improving model efficiency.
– Deployed the platform as a containerized service using Docker.
Technical Skills
Languages: Python, JavaScript, TypeScript, Java, SQL
Backend Development: FastAPI, Django, Flask, SQLAlchemy, Alembic, WebSockets, JWT
Frontend Development: Next.js, React, Tailwind CSS, HTML, CSS
AI & ML: TensorFlow, Keras, scikit-learn, Optuna, Pandas, Plotly
Databases: PostgreSQL, MongoDB, SQLite, Redis
DevOps & Cloud: Docker, Nginx, Gunicorn, AWS (EC2, S3), GitHub Actions (CI/CD), Linux
Developer Tools: Git, Postman, Vite
