Joseph Redd Consulting

Case study 9 Jun 2026

Project name

Not applicable

 

Client and industry

Don C., FinTech

 

The problem

  • 10 hours wasted weekly running the same day trading algorithm
  • $2k – $10k weekly opportunity earnings lost due to productivity inefficiencies
  • 10 hours wasted weekly trying to get Gemini to create and/or perfect an effective prompt

 

Hypothesis

This client wanted to create a prompt that would automate his day trading actions.

Specifically, he wanted to use AI to test his line of regression theory for maximizing opportunities for increasing and earning yields. Then, he wanted AI to create a “limiter” on when to take certain investment actions (e.g., selling options). Generally, he felt his way of going about utilizing AI seemed inefficient on the prompt engineering side, and ineffective on the day trading activity side.

I hypothesized that he needed an AI audit at minimum.

(An AI audit is a 90-minute deep dive, a thorough exploration of a client’s operations and current workflows. I use my audit framework to analyze where a client’s AI and apps level and/or use can be activated, improved, evolved, etc., to create the most business leverage. My AI audit framework usually produces a 5-10 document with findings and recommendations.)

 

The solution

  • Joseph Redd Consulting AI Audit
  • Joseph Redd Consulting AI Audit 60-Minute Walkthrough
  • Joseph Redd Consulting Implementation Roadmap Phase 1

Timeline: 90 minutes

 

Action

I expected to be able to suggest a different tool for this client to use, and then he be able to proceed on his own from there.

In actuality, he was using general AI tools—viz., Gemini and Grok—expecting super agent results. Realizing this, I explained a bit about the differences among general AI tools versus AI tools usable for project training purposes (e.g., ChatGPT, Claude, etc.). I didn’t need him to agree with me or change this tools; I just wanted him to be aware of the limitations of his preferred tools.

Regardless, I did aim to add an AI tool to his list of tools for consideration: Perplexity.

For example, the client explained that he wanted to access the past 6 months of trading data to test his day trading algorithm as theory. Gemini had informed him that data sets of that scale were not accessible by it for his intended purpose. (If the client engaged a similar attempt via Grok, he didn’t mention it to me.)

He seemed convinced and made note to try it out later.

As for perfecting his eventual prompt, the client wanted to pit AI tools against each other—viz., Grok, Gemini, and Claude—in effort to see which tool would produce the best prompt. I encouraged him to proceed, so that he can learn the differences and nuances among each tool. However, I discouraged him from getting carried away with the exercise thinking that it would get him to his actual goal faster—which, of course, is to increase yields. Specifically, I told him that the point isn’t to produce the best prompt for Perplexity to use. The purpose of the effort is to generate an effective prompt sufficient in description to allow Perplexity to test his theory. The client seemed convinced at that thought. As such, he also made note to try Claude (for perfecting his prompt for eventual insertion into Perplexity).

As a final thought, I advised the client that I’m more than happy and qualified to audit his current AI need and use; however, at some point in the near future, he may want and/or need to speak with a consultant who specializes in FinTech and that I can refer him to someone.

With that, the AI audit largely was completed.

 

The actual result

  • New free tier client accounts created for Claude and Perplexity
  • 1 bespoke prompt engineered by Claude for use in Perplexity
  • An AI audit 1-page document with findings and recommendations

 

Client quote

“i [sic] just invented a new indicator.  Google.ai says it has an edge.” (Don C., Day Trader (solopreneur))

Insight

This client was so excited about the consultation that I gave him that he stayed up late that night to implement my recommendations and suggestions. He followed-up with me the next day. Sheer excitement on his part.

I learned several things helping this client:

  • There remains an opportunity for AI tool providers as well as consultants to educate folks on the intended purpose of any/each AI tool.
  • I can consult any sole proprietor, solopreneur, or small/large company via my AI Audit Framework.
  • Gemini and Grok did a great job notifying the client of limitations, but it failed to really assist him in solving his actual problem.

While I don’t specialize in FinTech, I aim to follow up with this client if for no other reason but to see whether I can’t help him increase his yields by referring him to a specialist from my community of AI consultants and app builders.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top