639 - How To Stop ChatGPT From Lying To You (AI PhD Reveals The Fix) With Garima Agrawal Paul Higgins podcast

Content provided by Paul Higgins. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by Paul Higgins or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://podcastplayer.com/legal.

Paul Higgins Podcast »
639 - How to Stop ChatGPT from Lying to You (AI PhD Reveals the Fix) with Garima Agrawal

4d ago 41:03

MP3•Episode home

Why you should listen

Garima Agrawal bridges the gap between PhD-level AI research and practical consulting applications—she understands both the Formula One engine and how to help you drive the car better.
Learn the layered prompting technique that eliminates most hallucinations and turns ChatGPT into a thinking partner rather than a content generator you can't trust.
Discover which LLM to use for specific tasks—ChatGPT for analysis, Claude for coding, Gemini for technical research, and Perplexity for citations—so you stop wasting time with the wrong tool.

You're using ChatGPT every day, but deep down you don't fully trust it. You've been burned by hallucinations. You watch it confidently give you wrong information, and you're left wondering which parts are reliable and which parts are fiction. You know AI should be making you more productive, but instead you're second-guessing every output and spending more time fact-checking than creating. Meanwhile, you see other consultants talking about Claude, Gemini, Perplexity—and you're not sure if you should be using those instead or if ChatGPT is fine. You're stuck in this loop of knowing AI is powerful but not knowing how to harness it properly. This week, I sat down with Garima Agrawal, who brings a rare combination: PhD-level AI research expertise and real-world consulting experience running her own firm. She breaks down the exact framework that eliminates most hallucinations, reveals which LLM actually performs best for specific consulting tasks, and shares the prompting mistakes that sabotage your results before you even start. If you're ready to stop fighting with AI and start using it as the leverage tool it should be, this conversation gives you the practical roadmap you need.

About Garima Agrawal

Garima Agrawal is the founder of HumaConn LLC, a consulting firm focused on empowering businesses and investors with tailored AI solutions. At HumaConn, she specializes in designing strategies for effective AI integration—helping clients enhance efficiency, streamline operations, and identify high-potential AI-driven opportunities that bridge innovation with real-world impact.

She also serves as Head of AI Models and Services at Minerva CQ, where she leads the development and deployment of real-time, goal-oriented AI systems that power agentic, human-in-the-loop assistance in customer contact centers—boosting engagement and operational performance.

Garima holds a Ph.D. in Artificial Intelligence from Arizona State University, where her research focused on making AI knowledge-aware—enabling large language models to reason reliably in specialized domains. Her core expertise includes reducing hallucinations, optimizing retrieval-augmented generation (RAG), and building cost-effective, trustworthy AI pipelines.

With over a decade of experience spanning software engineering, data science, and AI leadership, Garima bridges the gap between cutting-edge research and real-world application—helping organizations build AI that delivers meaningful, measurable results.

Resources and Links

Humaconn.com
Garima's LinkedIn profile
Garima’s Google Scholar profile
ChatGPT
Claude
Gemini
Perplexity
Previous episode: 638 - The Rapid ROI Methodology
Check out more episodes of the Paul Higgins Podcast
Subscribe to our YouTube channel: @PaulHigginsMentoring
Join our newsletter
Suggested resources

317 episodes