Skip to content
View VanshGupta18's full-sized avatar
1 follower · 3 following

Highlights

  • Pro

Block or report VanshGupta18

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
VanshGupta18/README.md

Vansh Gupta

AI Engineer · Production-Grade Intelligent Systems

Building intelligent systems that move from notebook to production — RAG pipelines, LLM agents, financial analytics, and NLP applications.

About Me

I'm an AI Engineer focused on building reliable, production-grade systems that turn large language models into practical tools. My work spans the full lifecycle — from designing retrieval pipelines and agentic workflows to deploying full-stack APIs and building domain-specific NLP applications.

Right now, I'm deep in earnings call intelligence and corporate compliance automation, two domains where AI can meaningfully reduce analyst workload and governance risk. I'm also actively exploring geopolitical risk modeling and domain-specific fine-tuning.

Tech Stack

Languages

Python

AI & ML

PyTorch TensorFlow LangChain Hugging Face Groq scikit-learn Pandas NumPy Jupyter

Cloud & DevOps

Docker

🏆 Hackathon Recognition

Award Event Organizer Year
🥈 Finalist Meta PyTorch OpenEnv Hackathon SST, Bangalore 2026
🥇 Innovation Award Agentic AI Hackathon Ulster University, UK 2025

Areas of Focus

🔬 Currently building

  • Pressure-aware multimodal earnings call intelligence (text + audio + interaction)
  • OpenEnv-compliant agentic compliance systems with RL and LLM training pipelines (SFT + GRPO)

🌱 Actively exploring

  • Geopolitical risk modeling using LLMs and structured data
  • Domain-specific LLM fine-tuning for finance and legal
  • Agentic AI patterns: planning, memory, multi-turn tool use

Get In Touch

I'm open to collaborations on AI engineering projects, especially in financial analytics, agentic systems, or applied NLP.

Building systems that think — one pipeline at a time.

Pinned Loading

  1. Forsyt Forsyt Public

    Forsyt is an AI-powered, real-time geopolitical risk intelligence platform built specifically for Indian financial markets. It transforms global and regional news into the localized "India AI-GPR I…

    1

  2. corporate-compliance-env corporate-compliance-env Public

    The first open-source RL training environment for corporate compliance — train agents to audit expense claims against a policy rulebook across 3 difficulty levels. Built for the Meta PyTorch OpenEn…

    Python

  3. earning-calls earning-calls Public

    Pressure-aware earnings call intelligence system — models text–audio divergence, Q&A interaction dynamics, and managerial behavior under analyst pressure to predict returns, volatility, and risk si…

    Python 1

  4. ECG-Multi-Label-Classification ECG-Multi-Label-Classification Public

    Graph Neural Network (GNN) approach for multi-label classification of clinical ECG data using PyTorch Geometric, NeuroKit2, and the PTB-XL dataset.

    Jupyter Notebook

  5. Gina_cfp Gina_cfp Public

    Chat with your data. GINA turns plain-English questions into grounded answers by running real SQL on your dataset — not guessing from model weights. Built with Next.js, Fastify, Groq, and pgvector.

    TypeScript