Who is Shravan Gupta?

Shravan Gupta is an AI & Machine Learning Engineer based in Mumbai, Maharashtra, India. He specializes in Generative AI, Large Language Models (LLMs), Autonomous Multi-Agent Networks, and Full-Stack Web Architecture.

What are Shravan Gupta's core technical skills and expertise?

Shravan Gupta's technical expertise spans Generative AI, Large Language Models, Autonomous Agents, LangGraph, Model Context Protocol (MCP), RAG Pipelines, Vector Databases, Computer Vision, Python, PyTorch, FastAPI, Next.js, React 19, and TypeScript.

What research papers has Shravan Gupta authored?

Shravan Gupta authored and presented two peer-reviewed academic papers: 'Toward a Sustainable Future for Generative AI Technologies' (researching data center telemetry and sustainable foundation model computing) and 'AI-Based Intelligent Examination Invigilation Using Face Recognition and Behaviour Analysis from CCTV Surveillance' (addressing computer vision and multi-object tracking).

What notable engineering projects has Shravan Gupta developed?

Key production platforms built by Shravan Gupta include ML Doctor (mldoct, an enterprise PyTorch AST linter, C++ GPU VRAM watchdog, and ML loss diagnostics package on PyPI), BrandStudio (an autonomous multimodal AI social publishing bus), Outskill (an autonomous AI curriculum synthesis engine), and Yellow Crest (an industrial PMC platform with 60fps web telemetry and hardened security).

Where is Shravan Gupta's official portfolio hosted?

Shravan Gupta's official engineering portfolio and research showcase is located at https://shravangupta.dev.

Shravansay hello

12•

AI Math Learning Assistant

SolvYaar

Web & LLM Inference

Overview & Purpose

Interactive mathematical tutoring assistant featuring Groq sub-second LLM inference, step-by-step problem explanations, and gamified on-chain learning achievements.

System Implementation

Socratic AI Tutoring in Real Time

SolvYaar transforms math homework from a source of frustration into an interactive discovery session. Rather than simply giving the final answer, the AI guides students through foundational principles with step-by-step Socratic hints.

Metrics & Architecture

Production Results

Groq LLM

Sub-second step-by-step mathematical reasoning

ULTRA-FAST TOKEN STREAMING GENERATING PEDAGOGICAL HINTS AND LATEX EQUATIONS IN REAL TIME.

On-Chain

Learning achievements & verifiable milestones

ENCOURAGES STUDENT MASTERY THROUGH TAMPER-PROOF MILESTONE BADGES AND PROGRESS TRACKING.

Technical Architecture

High-Throughput Inference Engine

Connects to Groq ultra-low-latency LPU endpoints to stream step-by-step mathematical derivations in under 200 milliseconds.

Key Technologies

Groq API / Llama 3 Fast Inference
Next.js & TypeScript Architecture
KaTeX Math Equation Rendering
Tailwind CSS Modern Learning UI

Mobile Top View

Interactive Features

Developer Focus

Pedagogical Hint Scaffolding

Multi-turn conversational guardrails that ensure the AI provides hints and diagnostic questions before revealing the complete solution.

Architecture Patterns

Step-by-step Socratic hints
LaTeX syntax parsing & error recovery
Gamified progress telemetry
Adaptive difficulty scaling