ajay_dev

@Ajay_dev

Ajay Sable

Builder of AgentKit — Open-source workflow layer that takes AI agents from 20%...
Mumbai github.com/Ajaysable123/AgentKit Joined April 2026
674 Points6 Badges2 Connections2 Followers2 Following

About

Senior Gen AI Developer building agentic AI systems in production. Creator of AgentKit — the open-source orchestration layer that cuts Claude Code costs by 70%. 3.5 years shipping multi-agent pipeline

Top Skills

Vibe codingLLMAgentic AiRAGAi toolsPythonMulti Agent System

Language & Tools

ython · Node.js · Shell/Bash · SQLite · YAML · Claude Code · OpenCode · npm
(AgentKit's full stack — own what you actually built with)

Currently Exploring

Everything AI — agent frameworks, LLM orchestration, workflow automation, multi-agent coordination, and what it actually takes to make AI reliable in production.

Achievements

???? Built and open-sourced AgentKit — a workflow layer that boosts AI agent task completion from 20% to 80%
???? Published on npm with 127+ unique cloners in the first 14 days
✍️ Featured guest contributor at CoderLegion.com
???? Shipped 11 platform adapters, 5 intelligence layers, and a self-improving skill library

Fun Fact

My side project has more automated hooks than most production apps — and one of them automatically writes its own new skills from session experience.

Random Dev Quote

The model isn't the bottleneck. The missing structure is.
Joined: 3 weeks (since Apr 12)
Full Name: Ajay Sable
Headline: Builder of AgentKit — Open-source workflow layer that takes AI agents from 20% to 80% task completion
About: Senior Gen AI Developer building agentic AI systems in production. Creator of AgentKit — the open-source orchestration layer that cuts Claude Code costs by 70%. 3.5 years shipping multi-agent pipeline
Location: Mumbai
Website: github.com/Ajaysable123/AgentKit
Languges & Tools: ython · Node.js · Shell/Bash · SQLite · YAML · Claude Code · OpenCode · npm
(AgentKit's full stack — own what you actually built with)
Currently Exploring: Everything AI — agent frameworks, LLM orchestration, workflow automation, multi-agent coordination, and what it actually takes to make AI reliable in production.
Achievements: ???? Built and open-sourced AgentKit — a workflow layer that boosts AI agent task completion from 20% to 80%
???? Published on npm with 127+ unique cloners in the first 14 days
✍️ Featured guest contributor at CoderLegion.com
???? Shipped 11 platform adapters, 5 intelligence layers, and a self-improving skill library
Fun Fact: My side project has more automated hooks than most production apps — and one of them automatically writes its own new skills from session experience.
Random Dev Quote: The model isn't the bottleneck. The missing structure is.
Top Tags

Top Tags

Activity

User Activities

MayJunJulAugSepOctNovDecJanFebMarApr
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Less More
Contribution
chevron_left

Latest Jobs

View all jobs →

Active in these Groups:

AI
881 members
Open Source
831 members
Python Dev
451 members