RiskVision AI
An AI-powered financial risk analysis system designed to turn structured financial information into clearer risk signals and decision support.
View repository ↗AI Developer and Researcher From India
I'm Prem Narayan Sharma, a Computer Science student focused on building useful digital products, intelligent applications, and practical machine learning systems.
I enjoy the space where software engineering, data, and intelligent systems meet — turning an idea into something people can actually use.
My work spans full-stack applications, machine learning, generative AI, and experimentation. I care about clean interfaces, understandable systems, and shipping work that goes beyond a notebook or tutorial.
I'm currently building AI-powered applications and exploring modern ways of creating intelligent software while continuing to strengthen my software engineering foundations through hands-on work.
A selection of applications, experiments, and systems I've built.
An AI-powered financial risk analysis system designed to turn structured financial information into clearer risk signals and decision support.
View repository ↗An enterprise-oriented knowledge assistant for interacting with organizational information through natural language.
View repository ↗An AI/ML-powered workspace combining retrieval-augmented generation with a supervisor-driven workflow for retrieving papers, analyzing technical content, generating grounded summaries, and answering document-based questions with page-level citations.
View repository ↗A reinforcement learning project exploring Deep Q-Networks for safer overtaking decisions in a simulated highway environment.
View repository ↗A comparative Neural Style Transfer application exploring the trade-off between optimization-based quality and fast feed-forward inference.
View repository ↗An applied machine learning project for aircraft fuel consumption prediction, structured as a deployable application with training, backend, frontend, and containerization layers.
View repository ↗A full-stack job portal application focused on connecting candidates and employers through a practical recruitment workflow.
View repository ↗A practical toolkit for building, experimenting, and shipping.
Python · Java · JavaScript · TypeScript
Scikit-learn · TensorFlow · Keras · PyTorch · Transformers · LangChain · LangGraph
React · Next.js · Node.js · Express · FastAPI
NumPy · Pandas · Matplotlib · MongoDB · PostgreSQL · MySQL
Git · GitHub · Docker · AWS · Vercel · Linux
Product thinking · experimentation · clean implementation · continuous learning