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

01•

Omni-Channel Generative AI Studio, Social Media NLP & Automated Cross-Platform Publishing Bus

BrandStudio

Multimodal AI & Autonomous Agents

Overview & Purpose

An autonomous omni-channel social media AI platform—engineered like an AI-native Buffer. Combines brand-conditioned diffusion for photos and reels, multi-agent LLM copy & semantic hashtag ranking, acoustic & visual copyright verification with automated auto-fix, and synchronized scheduled dispatch across Instagram, LinkedIn, X, YouTube, and Facebook.

Visit Live Platform↗

System Implementation

Architecting an Autonomous Multi-Platform Social AI System

BrandStudio unifies omni-channel social management with autonomous multimodal generative AI and pre-flight content safety. Rather than juggling fragmented tools for graphic design, video editing, copywriting, copyright validation, and multi-network scheduling, companies configure their brand kit—logo, slogan, color schemes, and voice guidelines. The system coordinates specialized AI agents to generate brand-aligned photos and video reels, synthesize platform-tailored copy with semantic hashtag ranking, run pre-publishing acoustic copyright audits with automated audio auto-fix, and dispatch scheduled posts across Instagram, LinkedIn, X, YouTube, and Facebook either simultaneously or on granular account timelines.

Metrics & Architecture

Production Results

Multimodal GenAI

Diffusion synthesis & automated multi-aspect rendering across all socials

LATENT DIFFUSION AND GENERATIVE VIDEO PIPELINES PRODUCING ON-BRAND PHOTOREALISTIC ASSETS, CAROUSELS, AND SHORTS ADAPTED FOR INSTAGRAM, LINKEDIN, X, YOUTUBE & FACEBOOK.

Adaptive NLP & Compliance

Channel-tuned copy, semantic hashtag graphs & acoustic copyright auto-fix

MULTI-AGENT LLM ORCHESTRATION CRAFTING PLATFORM-SPECIFIC TONE AND HASHTAG GRAPHS WITH SPECTRAL AUDIO FINGERPRINTING TO DETECT AND AUTO-REPLACE COPYRIGHTED TRACKS.

Technical Architecture

Distributed Multimodal Agent Architecture

Coordinates multi-agent generative workflows, cross-platform media transcoding, automated compliance auditing, and asynchronous queue dispatch.

Key Technologies

Multimodal Latent Diffusion & Generative Video Synthesis (ComfyUI / Stable Video)
Platform-Adaptive LLM Copy & Semantic Hashtag Clustering (Claude 3.5 / Llama 3 70B)
Cross-Platform Acoustic Spectral Fingerprinting & Copyright Auto-Fix Pipeline
Dynamic Brand Kit Conditioning (Vector Encodings for Logos, Slogans & Hex Palettes)
Omni-Channel Social API Gateway (Instagram Graph, LinkedIn, X, YouTube, Facebook)
Distributed Redis BullMQ Scheduler with Exponential Backoff & Webhook Retries

Mobile Top View

Interactive Features

Developer Focus

Universal Content Safety & Cross-Platform Orchestration

Guarantees strict brand visual DNA and automated copyright compliance across every connected social media channel.

Architecture Patterns

Omni-channel simultaneous broadcast vs. platform-isolated scheduling
Automated multi-ratio media rendering (1:1 Square Feed, 9:16 Vertical Reels, 16:9 Landscape)
Pre-flight acoustic copyright scanner with automated royalty-free soundtrack replacement
Dynamic semantic hashtag indexing optimized against platform-specific shadowbans