Speaking

Talks for the audiences AI keeps misunderstanding.

I run a 200-year-old wedding venue and five AI products I built on top of running it, all shipped this year. The talks below come from production systems with the failures and trade-offs included. I speak to AI engineering audiences, IT and SaaS leaders, founders and investors, and the wedding and hospitality industry. Booking inquiries: info@isadoraandco.com.

From the course

A clip from my CEO course on AI.

A short excerpt from the course I built to introduce CEOs to AI. The same operator point of view I bring to a stage, compressed: what AI is actually doing under the surface, where it quietly works against you, and the parts leadership tends to get wrong before they get it right.

What I speak about

Three pillars.

Pillar one

Operator as builder.

I still run my own venue, and built every product I speak about on top of running it. These talks are about what changes when the person designing the AI is also the one it has to serve every day. That gap is something horizontal AI cannot replicate.

Pillar two

Voice and trust in AI agents.

How to make an AI sound like a particular business in production, not just at demo time. These talks go below the surface-level tone instructions that always collapse, into the architecture that actually holds. With real prompts.

Pillar three

Building for emotional industries.

Weddings, missing persons, mental wellness, personal AI. The audiences other AI products keep getting wrong. What the design discipline looks like when the user is in a heightened state and the standard playbook actively hurts them.

Featured talks

Six signature talks.

Each talk flexes between 20 and 60 minutes and adapts to keynote, deep-dive, panel, AMA, or podcast format. Tap any talk to see what it actually covers.

Founders · Marketing · Product · Agencies

30 min · standard

It Speaks to Them As They Wrote It

I own a wedding venue, and to run it I built a software suite around an idea I haven’t seen anyone else ship in production. A closed-loop intelligence system that listens to what couples ask and write about us, cross-references those signals with external data like economic confidence, seasonality and regional travel trends, and quietly feeds the resulting language, imagery and answers back into our website, marketing and client responses. Couples land on our site and read words that sound like their own. Confident answers to questions they hadn’t fully formed yet. It speaks to them as they wrote it.

AI engineers · Product · Designers · Founders

30 to 45 min · standard or keynote

Forbidden Words, Forced Pauses, Locked Modes

I build software for three audiences who share one thing: a generic AI response will hurt them. Couples planning their wedding, in the most emotionally loaded purchase of their lives. Families of missing people, in active grief, vulnerable to anything that sounds like false hope. And anyone capturing their own worst moments in a private journal, where a chipper AI cheerleader does real damage. The interesting design work in each of those products isn’t what the AI is allowed to say. It’s the words that are forbidden, the pauses that are forced, and the protected modes the system locks itself into when it notices the user isn’t okay.

Founders · Operators · Hospitality · Service-business leaders

Keynote · 30 to 45 min

AI for Relationship-Driven Businesses

Most AI sales pitches assume your business is transactional. Requests in, automation out, labour saved. That logic works for shipping logistics. It collapses for businesses where the relationship is the product: weddings, therapy, high-end hospitality, education. Anything where the customer is buying you. This talk is about that difference, with specific examples of where AI deepened the relationship at my own venue and where I had to actively rip it back out.

Founders · Investors · Product strategists

30 min · strategy talk

The Case for Vertical AI

Horizontal AI products get the press. Vertical AI products take the market. This is the operator’s case for going deep into one industry instead of broad across many. I run five products in five different verticals on the same technical spine. The ones that work and the ones that don’t differ by exactly one thing: how deeply they encode the actual workflow.

IT leaders · SaaS · Founders · Operators

Standard 45 min · or AMA

Replacing My Inbox With an Agent

I own a wedding venue. Earlier this year I retired the off-the-shelf tools that had been running my inbox and built my own Claude-powered agent to take over the most emotionally loaded email job in my business — talking to couples planning their wedding day. This is the from-the-trenches report. What it actually saved in hours, money and bookings. What it cost to run. The worst thing it ever did, and the moment I almost shut it down. The trust framework I had to build before it could talk to vulnerable customers. And the parts I would do completely differently if I started over today. For people being told to “do AI” by leadership, this is the version with the marketing stripped off.

Wedding & hospitality · Hospitality SaaS · Legacy-industry founders

Keynote · 30 to 45 min

The Wedding Industry Is Ten Years Behind

Most wedding industry software is a customised CRM with worse UX. Meanwhile the customer base, which is to say couples, has quietly become the most digitally fluent and brand-aware group of consumers in any luxury vertical. The gap between what the industry sells and what the customer is reading on Instagram is measured in years, and it is widening. This talk is the operator’s case for why we got here and what the next decade of luxury hospitality tech needs to look like.

Full catalogue

Eight more, on request.

Talks I give less often, sized for narrower audiences or specific formats. Email for the full abstract.

AI engineers · Wellness & mental-health tech

AI for the Person at Their Lowest

When someone tells your AI they hate themselves, the first thing it should do is not list crisis resources. Designing AI that knows the difference between a hard day and a real crisis, and responds at the right depth.

AI engineers · Prompt engineers · Agencies

The 4-Layer Prompt for Brand Voice

The four-layer architecture that gets a Claude agent to write as a specific brand, every time, including the edge cases the original prompt never covered. With real prompts and the multi-tenant pattern.

Solo founders · Engineers · "AI coding in real life"

Working With Claude Code As Your Engineering Team

Five production codebases shipped solo in the last quarter, no engineering team. The patterns that survived contact with reality, the guardrails I added after losing data, and the moments to interrupt the model and write the spec yourself.

AI engineers · Product · Designers · Trust & safety

AI Disclosure That Can’t Be A/B Tested Away

If your AI talks to your customers, the disclosure has to be structural, not configurable. A small talk about a specific design decision that generalises to every honesty commitment in your product.

AI engineers · Architects · Agency builders

Reactive Intelligence as a Design Pattern

The closed-loop pattern (ingest, classify, enrich, reflect) abstracted from five production systems. The schema shapes that recur, the failure modes, and the question that decides whether you can apply it to your product.

AI engineers · Wearables · Personal AI

Building a Wearable + AI Loop That Actually Works

Omi → webhook → Reid. What ambient capture actually means for prompt design, latency, hallucination, and trust. Including the line between “Reid heard this” and “Reid will mention this.”

Wedding & hospitality · Marketing · Sales operators

Why Your Booking Funnel Is Leaking

Most luxury venues lose 40–60% of qualified inquiries. The operator’s guide to fixing it, using my own venue’s data. AI is part of the answer; rewriting the inquiry experience itself is most of it.

Founders · Civic tech · Ethics & responsibility

Free Public Utilities Built With AI

Threadline costs me $500 a month to run, helps families I will never meet, and is not a startup. The case for non-commercial AI products, and the design freedom of removing the business model.

What I bring

Operator credibility, with the receipts.

350+

Weddings hosted at Rixey

Production AI products

12 wk

End-to-end build window

Operator, no engineering team

Booking inquiries

Direct, no agency, no gatekeeper.

Email with the audience, format, date, and any context on what you want the talk to do. I respond personally and will tell you straight if a different talk would land better than the one you asked for.

Email LinkedIn

Press kit & bios →← Back to Isadora & Co