Browse categories
Explore
Fiverr Pro
English
$
USD
I will perform QA testing for your ai agent reliability
Mazu
About this gig
AI agents are complex. They get stuck in loops, misuse tools,
and hallucinate.
I test your AI agent's reliability, tool use, and
decision-making.
WHAT I TEST:
Core Logic Does it follow instructions?
Tool Calling Does it use APIs correctly?
Memory Does it remember context?
Edge Cases How does it handle weird inputs?
Safety Does it avoid harmful actions?
PACKAGES:
BASIC ($100) Core loop & edge case testing + quick report
STANDARD ($250) Full tool, memory & multi-turn testing +
detailed report
PREMIUM ($400) Complete audit, CI/CD eval setup, re-test &
optimization plan
️ TOOLS: LangSmith, LangFuse, DeepEval, RAGAS, Custom
Python scripts
️ WHY THIS MATTERS:
Broken agents destroy user trust
Infinite loops drain your API budget
Tool misuse causes real-world errors
Investors require agent reliability proof
Message me before ordering for a free initial assessment.
Limited: 15 clients per month.
Get to know Mazu
Mazu
"I Protect Your AI From Security Risks, Bias Compliance Failures"
- FromPakistan
- Member sinceNov 2025
- Avg. response time1 hour
Languages
Urdu, English
I'm an AI Security & Trust Engineer passionate about making
AI safe, reliable, and legally compliant. I help businesses
deploy production-ready AI systems that don't hallucinate,
leak data, or fail audits.
My work spans AI red teaming, building LLM guardrails,
evaluating agents, developing custom RAG pipelines, and
creating EU AI Act/NIST compliance documentation.
I use tools like LangChain, RAGAS, PyRIT, Pinecone, and
LlamaGuard to ensure your AI is secure, accurate, and ready
for enterprise.
Message me to discuss your project.
My Portfolio
FAQ
What AI agent frameworks do you test?
LangChain, LangGraph, CrewAI, AutoGen, Semantic Kernel, and custom Python implementations.
Do you test multi-agent systems?
Yes. The Standard and Premium packages include testing for multi-agent coordination and communication.
How do you test tool calling?
I simulate various user intents to verify the agent selects the right tools, formats parameters correctly, and handles API errors gracefully.
What is "CI/CD eval setup" in the Premium package?
I set up an automated testing pipeline (using GitHub Actions + DeepEval/RAGAS) that runs every time you update your agent's code.
Can you test agents that use multiple LLMs?
Yes. I can evaluate agents that switch between GPT-4, Claude, Gemini, or open-source models.
What deliverables do I get?
A professional PDF report with test results, failure analysis, CVSS-like severity scores, and actionable fix recommendations.
Do you fix the issues you find?
Standard includes detailed fix recommendations. Premium includes hands-on optimization support and a re-test.
How long does the testing take?
Basic: 3 days. Standard: 4 days. Premium: 6 days. I start within 24 hours of receiving access.
Do you test for prompt injection in agents?
Yes, but I recommend my Gig 1 (AI Red Teaming) for a comprehensive security audit. This gig focuses on reliability and performance.
What do you need to start?
Access to the agent (URL, API, or repo), a description of its goals, and a list of tools it can use.
