Skip to main contentHomepage

Hi, I'm Yash. I build
AI Agents
that automate complex workflows 24/7.

Yash Sarnaik - AI & ML DeveloperYash Sarnaik - AI & ML Developer
Services & Capabilities

What I Build For Businesses

I engineer practical, cost-effective AI solutions tailored to solve specific business bottlenecks — built for production reliability, not just demos.

Have a custom AI requirement?

Let's evaluate if an AI agent, RAG pipeline, or custom model makes sense for your product.

Book a Discovery Call
Featured Work

Selected Case Studies & Systems

Production AI pipelines, multi-model architectures, and diagnostic models built for real-world impact.

How We Work Together

Simple, Transparent 3-Step Process

From initial concept to production-ready AI — here is how I take your idea from whiteboard to reality without technical friction.

15-30 Min Call

Discovery

Discovery

Discovery & AI Feasibility
7–10 Days

Prototype

Prototype

Rapid Working Prototype
Production Rollout

Deployment

Deployment

Production Deployment & Guardrails
Background & Details

About Me

A snapshot of my location, favorite stack, tools, and social channels.

Tech Stack

GitHubAnacondaStreamlitPyTorchGoogle ColabPostmanDockerJupyterPythonPostgreSQLPower BISupabaseFirebase
GitHubAnacondaStreamlitPyTorchGoogle ColabPostmanDockerJupyterPythonPostgreSQLPower BISupabaseFirebase
scikit-learnMySQLpandasGitNumPyVisual Studio CodeCloudflareFastAPIFlaskTableauTensorFlowPydantic
scikit-learnMySQLpandasGitNumPyVisual Studio CodeCloudflareFastAPIFlaskTableauTensorFlowPydantic

Fav. Web Framework

FastAPI

Fav. Automation Tool

Technical Foundation

Production Stack & Capabilities

The modern AI and infrastructure stack I use to build fast, scalable, and reliable systems.

LLM Orchestration & Agents

Autonomous multi-agent execution, state machines, and reasoning chains.

LangChain & LangGraphState machines, structured chains & memory
CrewAI Multi-AgentsAutonomous role-playing agent networks
Hugging FaceModel fine-tuning, embeddings & transformers
Python
PythonCore AI engine, async pipelines & scripting

Unlocks: Autonomous multi-agent swarms, complex reasoning workflows, and custom tool calling.

Vector Search & Knowledge RAG

Semantic vector databases, hybrid retrieval, and knowledge indices.

Pinecone & ChromaDBHigh-dimensional semantic search indexing
PostgreSQL
PostgreSQL & pgvectorRelational storage with native vector extensions
pandas
NumPy & PandasData cleaning, matrix manipulation & preprocessing
MySQL
SQL & Database OptimizationComplex queries, schema design & caching

Unlocks: Sub-second document search, private QA with zero hallucinations, and hybrid ranking.

Backend & High-Throughput APIs

Production-ready AI microservices, streaming endpoints, and dashboards.

FastAPI
FastAPI & Async PythonHigh-performance REST & WebSocket streaming APIs
Crawl4AI & PlaywrightDeep web scraping for custom LLM RAG pipelines
Streamlit
Streamlit & GradioRapid model demonstrations & internal tooling
Jupyter
Jupyter & ColabExploratory data analysis & model benchmarking

Unlocks: Low-latency token streaming, secure client authentication, and rapid UI prototyping.

Infrastructure & Automation

Docker containers, cloud deployments, and workflow automation.

Docker
Docker ContainersIsolated AI microservices & reproducible builds
Amazon
AWS & GCP CloudCloud Run, EC2, S3 storage & serverless hosting
n8n AutomationVisual trigger workflows, CRM integrations & bots
Git
Git & GitHub CI/CDAutomated testing, continuous deployment & versioning

Unlocks: Predictable cloud infrastructure, automated syncs, and 99.9% production availability.

What people say about me

Aniket Mhalungekar

Aniket Mhalungekar

Product Engineer Fullstack

The autonomous coding agent setup completely changed how I approach development. With background task execution, automatic PR creation, and preview deployments, a lot of the manual intervention was eliminated and it saved me a ton of time.

Open for Projects & AI Consulting

Ready to Turn Your AI Vision into Production?

Whether you need autonomous AI agents, NL2SQL pipelines, on-device Edge AI, or WhatsApp automations — let's evaluate your project on a 15-minute Google Meet call.