👋 Hi, I'm Krishil AgrawalAI ML Engineer &Agentic Developer.

Building multi-agent architectures, RAG pipelines, and end-to-end ML systems using LangChain, LangGraph, and PyTorch.

Based in Vadodara, India.

Hire Me as AI Engineer

View My Projects

About Me.

I engineer intelligence.

I specialize at the intersection of Machine Learning and autonomous Agentic workflows. From building robust LangChain orchestration engines to deploying low-latency PyTorch models, I bridge the gap between AI research and highly scalable, production-ready applications.

9.27

Current CGPA

3rd-year B.Tech CS Student consistently ranking among the top peers in quantitative engineering.

Based In

Vadodara, India

Available for Hire

Ready for full-time engineering roles & freelance AI architecture.

Download Resume

Experience &
Timeline.

AI Research Intern

05/2025 – 06/2025

CHARUSAT University, Gujarat — Deep Learning & Medical Imaging

✦
Building an AI-based system for affordable eye disease screening in rural India via retinal fundus image analysis.
✦
Designing a deep learning classification model using MobileNetV3Large with two-stage transfer learning to identify 4 conditions (CNV, DME, Drusen, Normal) using mixed precision.
✦
Engineering a comprehensive multi-metric evaluation framework producing confusion matrices, ROC/AUC curves, and per-class F1 scores.

B.Tech Computer Science

2023 – 2026

CHARUSAT University, Gujarat

9.27Current CGPA

Certifications & Accolades

✓
Semi-finalist in 4 HackathonsDemonstrating rapid prototyping & execution.
✓
Agent Skills with AnthropicPrompt engineering, tool use, Claude agents.
✓
Multi-Agent Systems with CrewAIEnd-to-end architecture and deployment.
✓
Gemini CLI Open-Source AgentsBuilding extensible AI pipelines.

Selected Work.

Production AI Systems & Research. Click any project to view architectural deep dives and model evaluations.

LangChain, LangGraph, Socket.IO, Next.js

AI-Canvas

Production-grade AI SaaS platform automating LinkedIn workflows using LangChain and real-time Socket streaming.

✦LangSmith Tracing, real-time sync

LLMs, Resume Intelligence, Next.js

TalentoAI

End-to-end AI-driven platform for interview preparation using deterministic LLM evaluation matrices.

✦500+ Reports Generated

RAG, Vector DB, DeepSeek, Next.js

DEPSTAR Admission Bot

A powerful Retrieval-Augmented Generation chatbot managing domain-specific university admission protocols.

✦Reduced query wait time by 90%

PyTorch, MobileNetV3Large, Transfer Learning

Retinal Screening DL

Deep learning model utilizing MobileNetV3 for identifying 4 retinal conditions from OCT scans with 96% accuracy.

✦96.4% Accuracy, Mixed Precision

Insights & Writings.

Breaking down complex AI concepts, real-world constraints, and rigorous engineering architecture.

FeaturedAgentic AISystem DesignGenAI / LLMs

MCP vs A2A — The 2 Protocols Every AI Developer Needs to Know

MCP connects your agent to tools. A2A connects your agent to other agents. Those are two very different problems — and confusing them will wreck your architecture before you write a single line of code.

12 min read• April 2026

SecurityDeep DiveAgentic AI

Why Agentic AI Is Still Broken: 5 Security Failures Killing Real Deployments

Agentic AI promises autonomy, but prompt injection, tool misuse, and broken trust chains are silently killing deployments. Here's what's really broken and how to fix it.

10 min read

GenAI / LLMsSystem Design

Scaling RAG Systems: From Naive Retrieval to Agentic Chunking

80% of RAG failures happen at the chunking layer, not the LLM. Here's how to move from fixed-size splitting to intelligent, context-aware chunking.

14 min read

GenAI / LLMsSystem DesignAgentic AI

GraphRAG vs Vector RAG — When Relationships Beat Similarity

GraphRAG beats Vector RAG in 4 specific scenarios. Learn when entity relationships outperform semantic similarity — with diagrams, examples, and code.

12 min read

Stack & Toolkit

LangChain

PyTorch

TensorFlow

HuggingFace

n8n

CrewAI

Next.js

PostgreSQL

MongoDB

Prisma

Supabase

Docker