I will build ai agents and workflow automation with claude code


Über diesen Service
Need an AI agent that actually finishes tasks - not a chatbot that just answers questions?
MS Data Science, IBM Data Science Professional. I build multi-step AI agents with Claude Code, Python and n8n: agents that remember context, call your tools and APIs, and run inside guardrails so they never act outside the limits you set.
Most AI agent gigs ship a prompt wrapped in a UI. I ship something that does the job end to end - qualifies a lead, updates your CRM, pulls a report, monitors a process - with clean logs so you can see exactly what it did and why.
What you get:
- Multi-step agent with short or long-term memory
- Tool use: APIs, databases, internal scripts, scheduled tasks
- Guardrails - confirmation before any data-changing action
- Source code, architecture notes, and a recorded walkthrough
- A revision pass to tune behavior against your real inputs
Proof: an n8n research agent I built auto-summarizes sources end to end - shipped and running (portfolio).
Stack: Claude Code, Python, n8n, LangChain, TypeScript.
Place your order - describe your workflow in the requirements form and I will get started. Prefer to scope first? Message me, I reply fast.
Lerne Nisar Khan kennen
AI Agent Chatbot and App Developer MS Data Science IBM Certified
- AusPakistan
- Mitglied seitDez. 2022
- ⌀ Antwortzeit1 Stunde
Sprachen
Urdu, Paschtunische Sprache, Englisch
Mein Portfolio
Meine weiteren Dienstleistungen im Bereich KI-Entwicklung
FAQ
What's the difference between an AI agent and a chatbot?
A chatbot answers messages; an agent completes a task — it plans steps, calls tools (your APIs, DB, calendar), checks its own work, and acts inside limits you define.
Will it hallucinate or act without approval?
No system is hallucination-proof, but I design for it: grounded inputs, confirmation before any data-changing action, logging on every step, and guardrails. You see exactly what it did and why.
Do I need my own API keys, and what about delivery and revisions?
You keep your own keys (ownership/billing) — typical small-workflow usage runs ~$5–$40/month; I help with setup. Delivery and revisions are listed per package (1/3/5 revisions, 4/6/10 days, extra-fast 48h add-on).
Where does it run after delivery, and what can it connect to?
Your server, a cloud account you own, or a documented container — never locked to me. It connects to the tools you already use through their official APIs; I confirm fit during scoping.
Have you built this before?
Yes — an n8n + MCP research agent that auto-summarizes Hacker News, and a RAG support agent that cut first-reply time from about 4 hours to roughly 90 seconds. Happy to walk through either.
What if it doesn't do what I expected?
That's what revisions are for — I tune against real inputs until it performs as scoped; anything outside scope we discuss openly first.
Ongoing support?
Yes — a monthly Care Plan (~$279) covers monitoring, small fixes, and incremental improvements. I usually propose it after first delivery.
You have no reviews yet — how do I know you'll deliver?
Fair question. My credentials are publicly verifiable (MS Data Science; IBM Data Science Professional) and every metric here is from a real named build. The Basic tier is a deliberately small, low-risk pilot - and Fiverr's resolution centre always protects your order.
How fast do you respond, and who owns the work and my data?
I reply within a few hours — my hours overlap UK/EU mornings and US-East afternoons, async-first via Fiverr. You own all source code and deliverables on delivery; I keep no copies; NDA available on request.
Can the agent read my company documents too?
Yes — if you want it to answer from PDFs, your website or a database, see my AI chatbot gig; the two combine well (chatbot answers, agent acts).

