Building intelligent systems that search, reason, automate, and ship.
I'm Adarsh Kushwah, an AI engineer and builder focused on AI agents, LLM applications, intelligent automation, search infrastructure, developer tools, and deep-tech systems.
I like building software that does more than respond to a request.
I build systems that can:
understand
↓
reason
↓
retrieve
↓
use tools
↓
execute
↓
learn from context
↓
produce useful results
My work sits at the intersection of:
AI × Software Engineering × Automation × Infrastructure × Developer Experience
I care about making AI systems useful, composable, observable, fast, and actually shippable.
| Area | What I work on |
|---|---|
| AI Agents | Autonomous workflows, tool use, planning, execution and agent orchestration |
| LLM Systems | Multi-model applications, routing, context engineering and AI APIs |
| Search Infrastructure | Web search, scraping, hybrid retrieval, ranking and research systems |
| Developer Tools | CLIs, SDKs, APIs, automation tools and coding infrastructure |
| AI Automation | Systems that connect models, APIs, tools and real-world workflows |
| Local-First AI | Self-hosted, BYOK and privacy-oriented AI applications |
| Full-Stack Engineering | Production interfaces, APIs, backend systems and developer platforms |
Open-source · Self-hosted · AI-native
Jiro is an AI-native search and scraping platform designed for developers and agents.
9 search engines
+
12 social platforms
+
hybrid search
+
structured extraction
+
MCP
+
AI research
+
official SDKs
↓
one developer platform
Built around
Python FastAPI MCP Search Scraping AI SDKs Docker WebSockets
Why it matters
Instead of building separate integrations for every search provider and scraping workflow, Jiro provides a unified developer interface for search, extraction and AI-powered research.
A machine-readable catalogue of free and free-tier infrastructure for building AI agents.
Current catalogue
384 resources
222 without a credit card
70 LLM APIs
64 MCP servers
82 agent tools
90 infrastructure services
78 agent skills
The project is designed around freshness rather than static lists.
Links are checked automatically, verification dates are tracked, deprecated services are preserved, and machine-readable datasets are exposed for both developers and AI systems.
Built with
TypeScript YAML GitHub Actions JSON GitHub Pages AI Agents MCP
→ Explore Free AI Agent Stack → Live Catalogue
AI
├── AI Agents
├── LLM Applications
├── Agent Orchestration
├── RAG
├── Context Engineering
├── Tool Calling
├── MCP
└── Multi-Model Systems
Engineering
├── Python
├── TypeScript / JavaScript
├── Rust
├── Go
├── React
├── Node.js
├── APIs
└── Distributed Systems
AI Infrastructure
├── Search
├── Web Scraping
├── Model Routing
├── Memory
├── Retrieval
├── Automation
└── Developer Platforms
Don't build another wrapper. Build the system underneath it.
I prefer software that can survive outside a screenshot.
The interesting part of AI isn't only generation.
It's what happens when a model can:
retrieve → reason → call → execute → verify
Good APIs, SDKs, abstractions and developer tools make every future project faster.
A powerful model with poor context can produce a weak system.
A well-designed context pipeline can make ordinary models dramatically more useful.
Prototype quickly.
Test against reality.
Remove unnecessary complexity.
Ship again.
I build projects primarily around:
AI Agents
AI Infrastructure
Search
Web Intelligence
Developer Tools
Automation
LLM Applications
Open Source
→ autonomous AI agents
→ AI search & research infrastructure
→ context engineering
→ multi-model routing
→ browser automation
→ local-first AI
→ developer infrastructure
→ intelligent automation
The exact projects evolve.
The underlying goal doesn't:
Make software more capable without making it unnecessarily complicated.
Adarsh Kushwah · DevAnimecx
AI Engineer · AI Agents · LLM Systems · Search Infrastructure · Developer Tools
Who is Adarsh Kushwah?
Adarsh Kushwah, known online as DevAnimecx, is an AI engineer and software developer focused on AI agents, LLM applications, search infrastructure, automation, developer tools, and open-source AI systems.
What does Adarsh Kushwah build?
He builds AI-native developer infrastructure, autonomous agent systems, search and scraping platforms, LLM applications, automation tools, and local-first software.
What is Jiro Search?
Jiro Search is an open-source, self-hosted search and scraping platform with multi-engine search, social scraping, hybrid retrieval, AI synthesis, MCP integration and developer SDKs.
What is Free AI Agent Stack?
Free AI Agent Stack is a continuously maintained catalogue of free AI APIs, MCP servers, agent frameworks, developer tools, infrastructure and agent skills, with machine-readable data and automated link verification.
What technologies does DevAnimecx use?
His technical work spans Python, TypeScript, JavaScript, Rust, Go, React, Node.js, APIs, AI agents, LLMs, MCP, search systems, web scraping, automation and AI infrastructure.
What is DevAnimecx interested in?
AI agents, context engineering, LLM infrastructure, intelligent search, autonomous software, developer experience, automation, local-first AI and open-source technology.