Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a custom ai agent with langchain and langgraph
Muhammad Saud
About this gig
A chatbot answers questions. An agent does work. It reads a request, uses your tools and APIs, and comes back with it done.
I work in Python with LangChain and LangGraph. My own product, SettlePost, runs live at logicstack.org/tools and syncs data between business systems, the same kind of problem an agent solves: connect systems, follow rules and handle failures.
Good fits for an agent:
- Sorting, drafting and routing emails or support requests
- Pulling data from your systems, checking rules, writing results back
- Research tasks that gather information from several sources into one summary
- Multi step workflows where a person approves first
What you get:
- An agent built around your exact workflow
- Connections to your tools and APIs
- Guardrails: spending limits, approval steps and full logs
- Full source code, plus a plain English guide on how it works and how to change the rules
- A recorded walkthrough of the agent running on your real task
Message me with the task you want automated. I will reply with how I would build it and which package fits. If an agent is the wrong tool, I will tell you, and a simple script or a chatbot might serve you better.
Get to know Muhammad Saud
Muhammad Saud
AI developer building chatbots and agents that do real work
- FromPakistan
- Member sinceFeb 2026
- Avg. response time2 hours
Languages
Urdu, English
I am a Python developer from Pakistan. I build AI chatbots trained on your own documents and AI agents that carry out real tasks with LangChain and LangGraph. I also run a live product of my own, SettlePost (logicstack.org/tools), which syncs data between Shopify and Xero on its own, so the software I deliver is built to run in the real world, not just in a demo. Message me before ordering and I will tell you honestly if your job fits, and which package you actually need.
My Portfolio
FAQ
What do I need to provide before we start?
A clear description of the task, and access details for the tools or APIs the agent must use. We map the workflow together first, in writing, before any code is written.
Who pays for the AI model usage?
You do. The agent runs on your own OpenAI or Anthropic account. I build in spending limits and logging so you always know what it costs.
Can the agent take actions, or only read information?
Both are possible. For anything that sends, deletes or charges money, I recommend an approval step where a person confirms first. We decide this together during the workflow mapping.
What if my workflow does not fit these packages?
Message me. Bigger or unusual builds get a custom offer with the scope and price fixed before work starts. No surprises later.
Do you maintain the agent after delivery?
Each package includes fix support for defects in what I delivered. Ongoing changes and new features can be booked as follow up orders.
