Aug 27, 2026 · 34m · no-priors
Rethinking Legacy Data Infrastructure with Eon Co-Founders Ofir Ehrlich and Gonen Stein
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
On this episode of No Priors, host Elad Gil sits down with Eon co-founders Ofir Ehrlich and Gonen Stein to discuss the evolution of enterprise cloud data architecture and security in the era of autonomous AI. The founders explain why proprietary enterprise data is the ultimate competitive moat and detail how modern data foundations must adapt to manage non-human agent risks.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 15.6% of the talking time here. How this is scored →
The hosts as informed peer 5.3 Guest teaching 5.0 Guest disagreement 1.7 The hosts pushing back 1.4
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 19:00 Reframing the future of dashboards
Ofir directly dismisses Elad's suggestion that dashboards will go away in the agentic era, asserting that dashboard monitoring will actually explode to track unmonitored non-human actors.
Hardest push from the hosts ▶ 6:12 Questioning the 'data is new oil' clicheElad pushes past vague data analogies to question what real-world buyers like Google actually plan to do with bankrupt airline datasets in reinforcement learning pipelines.
Biggest teaching moment ▶ 16:34 Non-human agents dropping production tablesGonen educates the audience and host on the emerging threat vector where AI agents with legitimate enterprise credentials execute destructive actions with extreme velocity.
The host holds their own ▶ 4:06 Naming Mercor in the Spirit Airlines data auctionElad demonstrates his deep venture and industry network by citing the undisclosed detail that startup Mercor was actively competing with Google in bankruptcy court for Spirit's data.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Eon's Cloud Data Foundation and Capabilities | 6 | 3 | 1 | 1 | Elad demonstrates insider knowledge of AI bankruptcy deals by noting Mercor was the other bidder against Google for Spirit Airlines data. The guests elaborate collaboratively on why enterprise data is replacing ephemeral compute as the ultimate moat. | |
| The Need for Real-World Enterprise Training Data | 7 | 4 | 2 | 3 | Elad challenges the overused 'data is the new oil' platitude by drilling into the technical utility of post-training, RL, and specific tooling like Applied Compute. Ofir explains how severe the shortage of non-synthetic real-world enterprise training sets actually is. | |
| Solving Data Silos and Security with Eon | 5 | 5 | 1 | 1 | Ofir acts out a detailed scenario explaining how internal corporate silos and security fears prevent data leaders from accessing enterprise databases. Elad concisely synthesizes Eon's value proposition into aggregation, masking, and model exposure. | |
| Cybersecurity Risks from Human to Non-Human Actors | 4 | 6 | 1 | 1 | Elad asks about sandbox-escaping models, prompting Gonen to educate on the transition from AWS-era human ransomware attacks to non-human AI agents possessing legitimate credentials dropping database tables. | |
| Non-Human Identities and the Democratization of AI Building | 5 | 6 | 4 | 2 | When Elad speculates that dashboards and human analytical interfaces might disappear in an agentic world, Ofir explicitly rejects the premise, arguing dashboards will proliferate to monitor out-of-control non-human identities. | |
| Rethinking the Data Plumbing for the Agentic Era | 4 | 6 | 2 | 1 | Elad inquires about the fate of existing ETL pipelines, and Ofir details why single-purpose historical tools like Fivetran and dbt are insufficient for the multi-contextual data ingestion required by autonomous agents. | |
| From Cloud Migration to AI Acceleration | 6 | 5 | 1 | 1 | Elad leverages the founders' previous experience building CloudEndure and leading AWS migrations to frame the shift to AI. The founders outline how forward-deployed engineering and private equity rollups are driving rapid enterprise AI transformation. |
Statements from this episode (12)
Ehrlich says proprietary data is the only real moat for AI
“When you can think about it, ah, models, ah, compute, everything is relatively ephemeral, ah, almost zero switching costs, and those are infrastructure, important parts of the infrastructure. For the industry. But if you're a company, whether you're a hotel ch…”
Ehrlich says Google bought bankrupt Spirit Airlines' data for $10 million
“Google buy something from the bankrupt Spirit Airlines. They didn't buy airplanes. They bought the data. They bought the data for ten million dollars, because they think it's very important in that perspective, they're using it to train models.”
Ehrlich says AI labs are trying to buy hedge fund data
“I'm hearing about, you know, labs going to Wall Street and trying to buy data from a hedge funds and try to understand how to map and analyze companies.”
Ehrlich says Spirit Airlines data trains AI on enterprise corporate hierarchies
“And I think that Spirit Airlines can be used both as an airline company, but also as a large enterprise, as a place where lots of people work, a lot of, you know, the hierarchy, middle management, top management, and workers working together.”
Ehrlich says misaligned incentives between data and business teams stall AI
“We understood that there's a problem with how this works because we have different incentives. You were tasked with doing that. I'm tasked with making sure my systems work, and I'm tasked with making sure that data is intact, no data is running away. I don't a…”
Stein says an unmapped AWS customer lost data to ransomware
“During our time at AWS, it was a very large customer that was impacted by ransomware. We thought that they were completely protected using our technology, the Disaster recovery service that we managed there, and we learned, unfortunately, that the customer tho…”
Ehrlich says enterprise leaders increasingly report data disruptions from AI agents
“Something that I noted is that like six months ago, no one would even discuss with me, but a few months ago, pretty much every person I meet, every a leader in a company tells me either they Are afraid of that happening to them, or it personally happened to th…”
Ehrlich says enterprises must assume breaches from approved internal AI agents
“We need to assume Breach whether it's malicious or not, and need to be able to handle it and act accordingly.”
Ehrlich says non-human identity is now a top-two enterprise cybersecurity issue
“How many cybersecurity companies you see in, in NHI, in non-human identity right now, an infinite amount. And there's a reason for that. It became a number one, number two problem right now.”
Ehrlich says legacy data tools cannot handle noisy AI agent data
“So you need tools that are able to both understand data from multiple locations, clean the noise, and make sure all of these data that's been created is actually usable. It doesn't apply with the old tools that were very, some of them were incredible. Five Tro…”
Ehrlich says the era of token-maxing is over as costs mount
“And we're not in the time of token maxing anymore. You know, trying to go to actually getting value for every token that we have, because it becomes more and more and more and more and more and more and more expensive.”
Ehrlich says forward-deployed engineers are essential for fast enterprise AI adoption
“Now it seems that you're coming to a large legacy enterprise. They really want to adopt the ad because they have to, the problem is they don't know how to do it. They understand that their processes are very long. There's something that takes a year or two or …”