I will fix and optimize your ai agent, chatbot or llm prompts

Is your AI agent or chatbot giving wrong answers, ignoring instructions, inventing information or calling the wrong tools? I will find the root cause and fix it.

I build and maintain LLM agents in production for an HR Tech platform (mainly Gemini, also OpenAI and Claude), including candidate screening agents and WhatsApp assistants. A big part of my job is designing prompts with strict context boundaries to reduce hallucinations and auditing AI-generated code for quality and security.

COMMON PROBLEMS I FIX

The agent misunderstands a specific instruction or user request

Hallucinations or answers outside your business info

Wrong tool or function calls in n8n or custom agents

Broken or inconsistent JSON / structured outputs

High token usage and slow responses

HOW I WORK

Step 1: I reproduce the issue with your real examples

Step 2: I find the root cause (prompt, context, memory, tools or code)

Step 3: I apply the fix and test it with multiple scenarios

Step 4: You get a short report of what changed and why

Please share access or exports of your prompts and workflow after ordering. Message me first if you are not sure which package fits.