I will build a custom ai agent using openai or claude API


About this gig
build AI agents that do real work, not just answer questions. That means agents that can look things up, call APIs, take actions in your tools, and make decisions based on your actual data.
Tech Stack:
- AI models: OpenAI, Claude, Gemini
- Tool calling / MCP, RAG (retrieval augmented generation)
- JavaScript, TypeScript
- Supabase (including pgvector for RAG)
What I Can Build:
- Support agents that handle inquiries end to end, with an optional human approval step before anything reaches the customer
- Document intelligence agents that answer questions from your files with citation backed answers
- Agents with tool calling that decide when to take an action, instead of relying on hardcoded rules
- Multi tenant systems with AI generated insights based on user specific data
I'm upfront about scope. If a project is really a large backend build rather than an AI agent, I'll tell you honestly instead of taking it on.
Message me before ordering with your use case and I'll recommend the right package.
Get to know Ivan B
Ivan B
- FromSerbia
- Member sinceJan 2024
- Avg. response time1 hour
- Last delivery2 months
Languages
English
My Portfolio
FAQ
Can you help refine my idea before we start building?
Yes. If you have a defined problem but haven't fully scoped the solution, message me and we'll figure out the right approach together before ordering.
What's the difference between a chatbot and an agent?
A chatbot mostly answers questions. An agent can take actions, like looking something up, updating a record, or deciding what to do next based on the situation.
Which AI model will you use?
Which AI model will you use?
Can the agent connect to my existing systems?
es, as long as there's an API available. Tell me what you're using and I'll confirm compatibility before we start.
Is this a good fit for voice agents?
This gig is focused on text based agents. Voice as a small add on to a text based system can work, but real time voice as the core deliverable is outside what I offer right now.
Do you provide support after delivery?
Yes, I include a short window after delivery for fixes and questions. Ongoing feature work is scoped separately once the agent is live.

