Case Study: TeachingTravel LLC — An AI-Personalized Newsletter Business
Making One-on-One Advice Scale
Using AI to give every client relevance a mass newsletter can’t, without the hours a real consultation would take
Skills: Seeing the Real Problem · Applying AI to a Real Constraint · Systems Thinking · Product Design · Content Strategy
TeachingTravel LLC is a New Jersey-registered business built to teach people how to travel extraordinarily well for a fraction of the retail cost, using credit card, hotel, and airline points and miles strategically. The proof of concept is personal: over the past four years, my family has flown and stayed in hotels worth $51,457.22 in retail value for under $7,900 combined in taxes, fees, and card costs. The business teaches others to do the same, through personalized consultations, educational content, and eventually a social media presence built to generate revenue through views and referrals. This case study focuses on one piece of that offering — the personalized newsletter tool — and how I’m using AI to build it.
The problem
Personalized travel and credit-card content is genuinely useful to people, but only if it’s actually relevant to them — their specific card portfolio, their travel interests, the deals that apply to them right now. Generic travel newsletters solve this by not solving it: they send the same content to everyone and hope some of it lands.
The reframe
Instead of choosing between “generic newsletter at scale” and “one-on-one advice that doesn’t scale,” the real opportunity was to make one-on-one relevance scale — using each client’s own data (their cards, their interests) as the input, and letting AI handle the matching and drafting that would otherwise take hours per client per issue.
The solution — what’s built
- An interactive web tool that generates each client’s newsletter directly from their intake data (currently Excel-based), matching relevant card news, deals, and reminders to their specific credit card portfolio and interests.
- A defined visual format for consistency and scannability: a yellow alert box for time-sensitive items, purple pill badges to tag which cards a piece of content applies to, italicized source attribution, and a purple call-to-action placed above the footer.
- An Airtable-based intake system, currently being refined — moving the consultation/newsletter opt-in to a single required multi-select field at the top of the form (rather than splitting it across two fields, which Airtable can’t cross-validate), with updated customer-facing framing and limited-time-free messaging to drive signups.
- An AI-generated spreadsheet that captures and maintains up-to-date information on the most popular credit cards on the market — including credits, benefits, and their associated timeframes (monthly, quarterly, semi-annual, annual, cardholder-year, or ongoing for as long as the card is held). The newsletter tool cross-checks each client’s card portfolio against this sheet and, based on proximity to the end of a given period, surfaces reminders to use relevant credits before they expire.
What’s next — the expansion
The current build is being extended so clients can:
- Update their own travel-interest and card-portfolio profiles directly, instead of going through manual intake
- Manage their own subscription preferences (frequency, topics)
- Interact with a full business website, which I’m building with AI tools rather than a traditional page-builder or agency
I’m also building a series of AI-generated educational videos covering:
- Understanding Personal Credit
- Credit Card Fundamentals
- Earning Points and Miles
- Redeeming Points and Miles
- Tracking Benefits, Credits, and Progress
- Tips, Tricks, and Advanced Strategies
- Issuer Rules You Must Know to Get Approved
- Including Business Credit Cards in Your Strategy
Why this is the strongest AI portfolio piece I have
It’s not a demo or a tutorial project — it’s a live business with real constraints: real client data, a real intake pipeline, and real tooling limitations (like Airtable’s validation gaps) that had to be designed around rather than assumed away. It shows AI applied to an actual operational bottleneck, not applied to show that I can use AI.
What I’d do differently
If I were starting today, I’d bring AI in earlier — not to refine a solution, but to help define the problem itself. Looking back, most of my AI use on this project fell into three familiar categories: using it to sharpen the wording on an overview presentation explaining points and miles, using it to identify a system (Airtable) to build an intake form, and using it to tweak that form once built. What I initially skipped was using AI as a genuine thought partner to question the premise: should TeachingTravel even offer an overview presentation? Personalized consultations? I’d started from my own idea of the solution and used AI to execute it, rather than starting from the open problem and thinking it through together.
That’s a habit I’ve since changed — and it turned out not to be a new skill so much as a familiar one. I’ve spent my career bringing people to a shared understanding of a problem before deciding what to do about it; a whiteboard full of colleagues was usually how that happened. This time, AI is the one at the whiteboard instead.
Current stage
TeachingTravel is still in the workshopping phase — no paying clients yet, and no manual version of this newsletter was ever sent to compare against. The tool described above is built and functional; what’s still ahead is launching it to real clients. That’s worth saying plainly here rather than implying traction that doesn’t exist yet. The personal savings figure in the opening framing ($51,457.22 in travel for under $7,900) is the credibility number this case study leans on for now — client-based metrics (subscribers, time saved, engagement) will get added once there’s a client base to measure.