Agentic AI Developer building LLM apps, automation workflows, RAG systems, and tools that can reason, call APIs, and get useful work done.
$ cat profile.json
{
"name": "Abraham",
"role": "Agentic AI Developer",
"location": "India",
"focus": [
"LLM applications",
"AI agents",
"automation workflows",
"RAG and knowledge systems",
"developer tools"
],
"status": "building practical AI systems"
}I build AI-native software that turns a goal into a workflow: plan the steps, use the right tools, read and write context, call APIs, verify outputs, and keep humans in control where it matters.
My favorite work sits at the intersection of LLM reasoning and real execution: agents that can search, summarize, automate, transform data, assist developers, and connect scattered tools into one calm command surface.
agent_workflows/
multi-step assistants, tool calling, task planning, browser/API automation
rag_systems/
embeddings, vector search, source-grounded answers, private knowledge assistants
automation/
scripts, bots, scheduled jobs, internal tools, repetitive-work removal
developer_tools/
CLI utilities, AI coding helpers, prompt/eval loops, workflow runners
languages:
- Python
- JavaScript
- TypeScript
ai:
- LLM APIs
- tool/function calling
- embeddings
- RAG pipelines
- prompt engineering
- eval-driven iteration
backend:
- FastAPI
- Node.js
- REST APIs
- background jobs
- databases
automation:
- browser workflows
- API integrations
- data pipelines
- terminal-first tools$ run agent --goal "ship useful AI products"
[01] understand the workflow
[02] design the agent loop
[03] connect tools and memory
[04] test behavior with real tasks
[05] deploy, observe, improve
status: onlinegithub : github.com/abrahamwebdev
twitter : twitter.com/realAbraham_GT
linkedin : linkedin.com/in/abraham-george-thankachan
planning -> tools -> memory -> execution -> feedback -> better agents