AI agents & automation

Hire an AI agent developer

I build AI agents that do real work — autonomous tool-using agents, voice and calling agents, and multi-step workflows wired into your systems. I'm Bhuwanesh Sisodia, a full-stack and AI engineer based in India working remotely worldwide, building agents with tool calling, memory, and human-in-the-loop guardrails on OpenAI, Claude, and the Vercel AI SDK.

What I build

  • Tool-calling agents that read and write to your CRM, database, and internal APIs
  • Voice / calling agents for sales, qualification, and support (telephony + LLM)
  • Multi-step, stateful workflows orchestrated with LangGraph
  • Memory, retries, and human-in-the-loop approval steps
  • Guardrails, evals, and observability so the agent stays on the rails

Stack

  • OpenAI
  • Anthropic Claude
  • Vercel AI SDK
  • LangGraph
  • Twilio
  • MCP
  • Python
  • Next.js

Questions, answered

What can an AI agent actually do for my business?
Handle inbound and outbound calls, qualify and route leads, answer support questions over your docs, run repetitive multi-step workflows, and take actions in your tools — all with guardrails and an audit trail.
Do you build voice and calling agents?
Yes — agents that place and answer real phone calls for sales, qualification, and support, talk naturally, book slots, and write results back into your CRM through tool calls.
How do you stop an agent from going off the rails?
Scoped tools, structured outputs, validation, human-in-the-loop approvals for risky actions, and continuous evals — so behaviour is bounded and measurable.