Walking through the streets of New York City to the Everyday AI Conference, it was impossible to miss the Dataiku ads lighting up subway stations across town. Every few steps, a new display would flash: a reminder of how far Everyday AI has come — and how big this movement is becoming.

But as I found myself reflecting on throughout the day, and in countless conversations with other attendees, a consistent theme emerged:

It’s not about how shiny or powerful your AI platform is — it’s about the business value you can prove with it.

This was a thread that ran through every keynote, session, and networking break, and it’s one that resonates deeply with us at Archetype — because delivering measurable impact is what sets us apart.

A Strong Start: Mirrors, Metrics, and Morning Keynotes

After arriving at the venue around 8:30am, we were greeted by what Dataiku always does so well: immersive visuals. To even reach the check-in kiosks, you had to walk through a brilliant mirror tunnel filled with statistical symbols and designs — a perfect photo op (and one that our own resident PhD, Marty Gaupp, especially appreciated!).

After grabbing a quick team selfie and breakfast (shoutout to Dataiku’s food game — bao buns and pho bowls at a tech conference? Unheard of!), we made our way to the opening keynote.

Keeping Afloat in the AI Ocean

Jonathan Zittrain, George Bemis Professor of International Law at Harvard Law School, kicked off the day with a keynote titled "Keep Buoyant in the AI Ocean." This was hands-down one of the best talks I've attended on how machine learning actually works — both educational and hilariously relatable.

His closing thoughts were a perfect setup for the day’s big theme:

• Keep an inventory of AI/ML deployments, whether in-house or outsourced

• Mind the supply chain and its architecture

• Set your risk and experimentation dial

• Practice "aggressive agnosticism"

• Activate internal engineers and data scientists

In short: it’s not about chasing the newest tech. It’s about knowing what you have, understanding its risks, and focusing on real outcomes.

Unlocking the Agentic Future of AI

Next came the Dataiku keynote led by Florian Douetteau (CEO) and Clement Stenac (CTO), with a fantastic lineup of guest speakers from Merck, Prologis, FINRA, and more.

One very cool moment: the intro music for each speaker was generated live using AI!

The major announcement? Dataiku’s new capabilities around AI Agent Creation and Control:

• Enterprise Orchestration: Agents connected to all data and models

• Continuous Optimization: Agents monitored and optimized

• Fully Governed Creation: Guardrails for agent development and deployment

Loved Florian’s analogy here: "Carrot & Stick – Creating Agents & Controlling Agents."

Again — a powerful set of tools, but it all comes back to what business value they actually drive.

Some standout comments and moments from the keynote guests:

• Mark Booth, CIO of Perdue Farms:

You don't need an AI strategy; you need AI built into your business strategy.

• Luke Slotwinski, VP of Data & Analytics at Prologis, on their contract clause detector and the critical role of enablement.

• Christie Hampton from FINRA on building a full Center of Excellence to scale AI adoption across their organization.

A fun moment was seeing the results of a live poll on agent deployment:

• 50: Already deployed agents

• 74: Agents in design

• 168: Plans for future agents

• 87: No plans yet

This really shows where most firms are — and why choosing the right first agent use case is so important. (Exactly why we built our AI QuickStart & Strategy Workshop: helping clients identify that "low-hanging fruit" that’s practical, valuable, and immediately impactful.)

Afternoon Breakouts: Lessons from the Frontlines

After lunch and some great networking, the breakout sessions offered real-world stories of success — and struggle.

Here are some highlights:

• 5 Lessons Learned from Supporting LLM Use at Bread Financial (Mindy Billingham): Critical to align LLM deployments with business costs and goals early.

• Building Bridges & Setting Boundaries at Cox Automotive (Brett Carpenter): Scaling 100s of models safely at 70% faster time-to-value — but only by focusing on sustainable AI practices.

• Climbing the Ladder at Takeda (Maria Camila Marenco): Outstanding real-world case of quantifying ROI. Their automation saved 43,350 hours/year for patients and unlocked ~$8M/year in potential cost savings. This was a masterclass in tying technical wins back to business outcomes.

• Transforming Contracts at Eaton (Rama Padhye): Extracting metadata from 30,000+ old legal contracts with 94% accuracy using Dataiku + Generative AI.

• Scaling Laws Panel with Zeus and ExxonMobil (Joe DelPercio & Erwann Lemenager)

"Companies cannot be technology first; it has to be business led with a business sponsor."

Every session reinforced the same truth: technology is the enabler, but business outcomes are the goal.

Wrapping Up with Connections (and Cocktails)

We ended the day with a packed networking happy hour, where the conversations kept circling back to the same questions:

• How do we prove the ROI of what we’re building?

• How do we bridge the gap between innovation and the business side?

These are questions we at Archetype live and breathe every day. It's not enough to deliver AI — you have to deliver AI with impact.

And that's exactly what we help our clients do.

Can't wait for the follow-up conversations with all the amazing people we met — and even more excited to keep helping organizations realize the real value of Everyday AI!

Interested in learning how to align AI with your business goals? Let's connect.