Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a custom ai agent with claude or openai and tools
Hasham Vakani
Level 2
About this gig
I build AI agents that ship to production, not demos that break on the second user.
I have built and deployed multi-LLM systems on HomeScanr (live on the App Store and Play Store, Claude and OpenAI routed across 5 services on AWS) and integrated agentic workflows into an enterprise ATS platform (NestJS and TypeORM, multi-tenant SaaS).
What you get:
- Production-tested architecture, not a prototype
- Claude or OpenAI, or both, routed by task
- Tools and function calling done right, with proper error handling, retries and observability
- Memory that scales (database-backed, not in-memory hacks)
- Full source code and a deployment guide, so you own it
I use Claude Code daily on live client work. The agents I ship handle real users, not demo scripts.
Tech: Claude API, OpenAI API, LangChain, LangGraph, Next.js, Node.js, Python (FastAPI), TypeScript, PostgreSQL, pgvector, Pinecone.
Not sure which package fits? Send me what you have and I will honestly tell you which tier you actually need.
Get to know Hasham Vakani
Hasham Vakani
AI agents, RAG, automation and MCP, production builds not demos
Level 2
- FromPakistan
- Member sinceAug 2020
- Avg. response time1 hour
- Last delivery2 weeks
Languages
English, Spanish, French
I ship production AI for startups - agents, RAG, automation, and MCP servers - and I fix AI-built apps that broke in production.
8,700+ hours delivered. 100% satisfaction. Claude Code daily on live work.
What I build:
- AI agents (tools, memory, Claude/OpenAI)
- RAG chatbots that answer from your docs with citations
- AI automation with n8n or Make
- AI voice agents for inbound calls and bookings
- MCP servers and AI SaaS MVPs (Next.js, Stripe)
- Fixing broken AI apps (Lovable, Bolt, Cursor)
Tech: Next.js, TypeScript, Claude, OpenAI, AWS.
Message me your use case.
My Portfolio
Other AI Development Services I Offer
FAQ
Which LLM do you recommend — Claude or OpenAI?
Depends on your use case. Claude is stronger for nuanced reasoning, long context, and tool use. OpenAI is stronger for raw speed and the broadest ecosystem. I'll recommend based on your actual workload after you share the use case.
Can the agent integrate with our existing API/database?
Yes. Standard and Premium tiers include 3-5 integrations. Bring API docs to the kickoff call and I'll confirm the scope before we start.
Do you handle deployment?
Basic delivers source code only. Standard delivers source + a deployment guide. Premium includes me deploying it to AWS/Vercel/Railway end-to-end.
Do I get the source code?
Always. Every tier includes the full source code (GitHub repo or zip — your choice).
What if I need more tools after delivery?
Order an extra revision or contact me for a custom offer. I keep a flat rate for additional integrations on completed agents.
How do you handle API key security?
Env vars + secrets manager (AWS Secrets Manager / Vercel env / your platform). Never hardcoded, never in the repo, never in client-side code.
What's your turnaround for urgent jobs?
24-48 hr expedited delivery is available as an extra ($X premium) on Standard/Premium tiers — message me first to confirm capacity.
Do you sign NDAs?
Yes, before any code or sensitive info is shared.
Can you train the agent on our company knowledge base or documents?
Yes — Premium tier includes a RAG-augmented agent that retrieves from your knowledge base before responding. For Basic/Standard, add the "Vector database setup" gig extra ($100) and we can wire in document retrieval too.
What's the difference between this gig and a simple chatbot?
A chatbot answers questions from a fixed prompt. An agent has tools — it can call your APIs, query your database, take actions, route to other agents, and remember context across sessions. If you just need a Q&A bot over docs, look at my RAG Chatbot gig instead — it's a better fit and cheaper.
