与Eon联合创始人Ofir Ehrlich和Gonen Stein一起重新思考遗留数据基础设施

到目前为止，威胁主要来自人为因素。

Up until now, the concerns came from human threats.

现在情况加剧了，我们看到同样的威胁来自非人类行为体，这些智能体实际上拥有合法的访问权限和正当的权限。

What we're seeing now on steroids is that the same type of threat is coming from nonhuman actors, agents that essentially have legitimate access to the environment with legitimate permissions.

幸运的是，在检测和防护方面，方法非常相似，但发生的速度极快。

Fortunately for us, it's a very similar methodology in terms of detecting that and protecting against that, but the velocity of that happening is extreme.

想想那些非技术人员。

Think of the nontechnical people.

他们甚至对安全、合规或谁将使用这些数据毫无概念。

They're not even aware for things like security or compliance or who is gonna use this data.

也许他们正在构建的智能体正在使用其他智能体，而他们甚至不具备理解其含义的技术背景。

Maybe their agent that they are building are using other agents, and they're not technical to even understand what it means.

这就在组织内部形成了一整套行为体，不受组织规则约束，也不一定在组织内部运行，却处理着敏感数据。

It creates complete set of actors inside the organization, not bound by the rules of the organization and not necessarily running within the premises of the organization, but handling sensitive data.

组织内每个人都能成为构建者，这既是好事也是坏事。

It's a good thing and bad thing that everyone inside the organization can become builders.

我们生活在一个非常有趣的时代。

We live in very interesting times.

今天在NoPriars，我们邀请到了Eon的联合创始人Ofir Ehrlich和Gonen Stein。

Today on NoPriars, we're joined by Ofir Ehrlich and Gonen Stein, the cofounders of Eon.

Eon是一家专注于云备份和灾难恢复的服务公司，专为AI时代设计。

Eon is a cloud backup disaster recovery centric services designed for the AI era.

在这次讨论中，我们聊了数据、AI，为什么谷歌在精神航空破产时买下了它的数据，以及在AI时代真正管理和使用数据基础设施意味着什么。

In this discussion, we talk about data, AI, why Google bought out the data of Spirit Airlines out of bankruptcy, and what it means to really manage and use data infrastructure in the AI era.

Ofir，Eon，非常感谢你今天来参加NoPrize。

Ofir Eonen, thank you so much for joining me NoPrize today.

很高兴见到你。

It's great to see you.

当然。

Absolutely.

谢谢你的邀请。

Thanks for having us.

对。

Yeah.

那你们在Eon做的事情之一，或者，其实，你们能不能简单介绍一下Eon是做什么的？

So one thing that you guys are doing at Eon is or, actually, why don't you give a quick overview of Eon and what it does really quickly?

因为我觉得这样能为我们讨论AI、数据、模型和微调模型奠定背景。

Because I think that'll set the context for how we think about AI and data and models and fine tuning models.

我觉得整个技术栈是建立在不同类型的数据集之上的。

I think there's a whole stack that's built on top of different types of datasets.

所以也许我们可以从你们做的事情开始，然后我想我们会逐步讨论企业数据栈是如何变化的。

And so maybe we can start with what you all do, and then I think we'll kinda walk through, like, how the world is shifting relative to to the enterprise data stack.

对。

Yeah.

当然。

Sure.

简单来说，我们做的事情是创建了一个运行在云端的新的数据基础，我们提供多种功能，让客户能够首先在多个云服务商之间映射和分类他们的数据，识别他们拥有什么、数据在哪里、哪些是敏感的、哪些不是，等等。

So, what we do at at a high level is we've created a new data foundation that runs in the cloud and we provide multiple capabilities that allow customers to first map and classify their data across their environment, across multiple hyperscalers and identify what they have, where they have it, what's sensitive, not sensitive, and so on and so forth.

然后我们提供了一种能力，可以轻松地从所有这些不同的来源摄取数据，包括结构化数据和非结构化数据，进入这个数据基础。

Then we provide an ability to easily ingest that data from all these different sources, structured, unstructured data into this data foundation.

而这个数据基础提供了一种非常经济高效的方式来维护数据以进行保护和恢复，同时也能够理解数据。

And the data foundation then provides a very cost effective way of both maintaining the data for protection and recovery, but also makes sense of the data.

所以它让客户能够非常轻松地访问数据、查询数据、搜索数据，并在从各种来源摄取的数据之上应用他们的人工智能模型和大语言模型。

So it allows customers to very easily access it, query it, search through it, and apply their AI models and LLMs on top of that data that's ingested ingested from a variety of sources.

对。

Yeah.

我的感觉是，你们的起点其实真的是作为备份和数据恢复及保护服务。

And my sense is, I mean, your starting point was really as sort of backup and data recovery and protection service.

我觉得在这个过程中，你们意识到，如果你从备份的角度拥有所有这些数据，并且拥有客户所有的历史数据，你就可以开始将其用于有趣的应用程序领域。

And I think along the way, kinda realize if you have all this data from a backup perspective and you have all their customer history over all time, you can start using that for interesting application areas.

那么，你们看到客户在哪些方向上利用这些完整的数据历史，也就是你们所拥有或代表的数据？

What are what are some of those directions where you're seeing customers take this the the sort of full history of data that they that you all have or represent?

正如你提到的，当我们开始时，我说过，我这个人有点疯狂，在一个AI的世界里创办了一家非AI公司，而AI的顺风车变得极其强劲，确保数据成为组织最重要的事情。

So as you mentioned, when we started, I said, I had this crazy person starting a non AI company in an AI world, and the AI tailwind became absolutely insane and made sure that data becomes the most important thing that an organization have.

你想啊，模型、算力这些东西其实都是暂时的，几乎没什么切换成本。

When you can think about it, models, compute, everything is relatively femoral, almost zero switching costs.

虽然它们也是行业基础设施的重要组成部分。

And those are important parts of the infrastructure for the industry.

但不管你是酒店连锁、餐饮连锁还是科技公司，你最有价值的资产其实是你的数据。

But if you're a company, whether you're a hotel chain or you're a food chain, a technology company, it doesn't matter, the most valuable thing that you have is actually your data.

越来越多的公司开始意识到这一点。

And you see more and more companies finding this out.

你知道吗，就在两天前，谷歌从破产的精神航空公司买了点东西。

You know, just two days ago you saw Google buy something from the bankrupt Spirit Airlines.

他们买的不是飞机。

They didn't buy airplanes.

他们买的是数据。

They bought the data.

他们花了一千万美元买这些数据，因为他们觉得这非常重要。

They bought the data for $10,000,000 because they think it's very important in that perspective.

他们用这些数据来训练模型。

They're using it to train models.

我觉得还有传言说，破产竞标过程中，另一个出价方是Mercor，对吧。

I think the rumor too is that the other bidder on the dataset was Mercor, right, in terms of the bankruptcy bid process.

所以这挺有意思的。

And so it's interesting.

有好几家AI领域的公司都在竞标一家破产航空公司的企业数据，这确实很吸引人。

You had multiple different companies in the AI world bidding on a bankrupt airlines enterprise dataset, which is fascinating.

是的。

Yes.

那你觉得

Do do

你觉得以后我们会看到更多这样的情况吗？

you think we'll be seeing a lot more of that in the future?

就是说，你觉得我们基本上会看到这种从破产公司手里买数据的情况吗？

Like, do you think we're gonna basically be seeing these, like, out of bankruptcy data buys?

到目前为止，我们在多个应用场景中都看到了这种情况。

So for we've seen it for multiple use cases.

这正是它最酷的地方。

That's what's really cool about it.

而且你看，Merkur，还有其他公司都在不断尝试购买数据。

And you see, Merkur, you see other companies are continuously trying to already buy data.

如果你现在是科技数据公司的CEO，我可以告诉你，你经常会收到这样的询问：你愿意出售你的数据吗？

If you're a tech data CEO today, I can tell you that you constantly get questions, are you willing to sell your data?

我到处都能听到这种声音，而且这似乎会成为一个重要的趋势。

I hear it all over, and it seems that's going to be a significant trend as you go.

我听说有些实验室去华尔街，试图从对冲基金那里购买数据，想了解如何绘制和分析公司。

I'm hearing about labs going to Wall Street and trying to buy data from hedge funds and trying to understand how to map and analyze companies.

所以你会看到，公司多年来积累的数据，通常像磁带一样，放在架子上积灰。

So you see a data that was accrued throughout the years by companies, which was usually like tapes, it was usually sitting on a shelf collecting dust.

突然间，这些数据变得非常重要。

And all of a sudden this becomes very important.

而且你会发现，现在公司都意识到，首先，我今天拥有的、能让我与众不同的，就是我的数据。

And you see companies now realize that first, what I have today that differentiates me than anyone else is my data.

这些数据就是金子。

And this data is gold.

而且我确实可以利用它来为公司创造更多价值，在AI真正进入竞争舞台的时候继续发展我的业务。

And I can actually leverage that to get more value for my company and to continue building my business when AI is actually coming fighting the playinggrounds.

看起来，无论是大公司还是小公司，基本上都在同一条起跑线上。

It seems that everyone, even large and small companies, basically have the same playing field.

而如今公司真正拥有的优势，除了人才之外，就是他们积累的数据。

And the only real advantage that companies have today is, of course, their people, but also the data that they've approved.

因为那些酷炫的新工具，大家都能用。

Because everyone has access to all of those cool new tools.

对。

Yeah.

这已经变成了一种护城河。

It's become a moat.

我觉得，就这个而言，人们一直说数据是新石油，这话说了很久了，我一开始还有点怀疑。

I guess in terms of the I mean, people have been saying data is a new oil for a long time, and I was a little bit skeptical of that statement.

但现在我觉得，因为后训练和强化学习，还有像Applied Compute这样的公司开始提供这些服务，你可以针对特定数据集微调模型或开源模型。

But I feel like now what's happening is because of post training and reinforcement learning and, you know, there's companies like Applied Compute and others who starting to provide these sorts of services where you can fine tune models or open source models against specific datasets.

看起来大家都在为自己的用例优化这些东西。

Like, seems like people are trying to optimize these things for their own use cases.

比如像Spirit Airlines这样的公司，是不是为了建一个航空公司应用而做客户支持？

I guess in the case of something like Spirit Airlines, is it customer support for building, like, a airline app?

你觉得他们实际上会用这些信息做什么？

Like, what what do you think they're actually gonna do with this information?

还是说这是别的东西，比如内部文档？

Is it something else that's the internal documents?

我就是好奇，这到底是强化学习，还是说它是一个客户支持代理？

Like, I'm just sort of curious, like, what what is the the reinforcement learning or is it like a customer support agent?

对。

Yeah.

想想看，如果你今天想构建智能体，你不能只在实验室里做，需要用新数据、一些训练数据来训练它们。

Think if you're trying to build agents today and trying to you can't just build them a lab, need to train them on new data, on some training data.

而且很难找到非常好的数据集。

And it's very hard to find very good data sets.

你看，Harpie 前几天刚发布了一个法律数据集，但你找不到太多好的数据集，它们看起来不像真正的合成数据，能真正反映现实世界。

You see that Harpie just released a legal data set just a few days ago, but you don't find too many good data sets that don't look like really synthetic data, that can actually be used to really look like the real world.

我认为 Spirit Airlines 既可以作为一家航空公司，也可以作为一个大型企业，一个有很多人工作的地方，有中层管理、高层管理和员工一起协作。

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 in middle management, top management and workers working together.

而且，如果你看看外面有哪些公开数据集，其实有很多。

And, you know, if you're looking at what other public data sets do you have out there, There are a lot of those.

真的，我在和公司交流，问他们有什么样的数据。

Seriously, I'm speaking with companies asking what kind of data do you have?

你们用什么来训练？

What do you train on?

你会找到一些东西的。

You'll find stuff.

比如说，Angular的数据是公开的，人们实际上把它当作一家公司的真实数据来用，因为公司就是这样运作的。

For example, the Angular data is out there in public and people are actually using that as real data from a company, because of how a company works.

原因在于，很难找到能帮助你在真实世界中工作的数据。

And the reason is it's so very hard to find data that will help you to work like in the real world.

每当你看到有人构建代理或开发新应用时，大多数其实并不真正管用。

Anytime you see someone building an agent or building a new application, most of them don't really work.

你必须走向世界，真正与真实世界的公司互动，才能做出有意义的东西。

You have to go to the world, you have to actually interact with real world companies in order to really build something significant.

现在你可以通过接触客户来实现这一点，你可以购买数据并在内部进行训练。

Now you can do that when you go to customers, you can buy data and train in house.

所以当你首次发布产品时，每个新产品你都不必先与客户互动作为初次接触。

So when you first release your products, every new product that you have, you don't have to first interact with customers as your initial interaction.

所以我认为你会看到越来越多这样的情况，既通过创新的方式创建新的合成数据，也获取现有数据，无论是你的真实数据，还是某种大规模的数据，想想看，它包含敏感信息，比如个人身份信息、财务信息等等，然后在此基础上构建真实世界的东西。

So I think you're going to see more and more of that, both by creating new synthetic data in new innovative ways, in addition to getting existing data, whether it's your real data, whether it's somehow massive, think about it, it contains sensitive information like PII, financial information, so on and so forth, and actually be able to build real world stuff on top of that.

对。

Yeah.

Google 显然也在这块，他们在旅游领域已经深耕一段时间了。

And Google obviously is not they're in this travel space for a while.

对吧？

Right?

他们非常需要这类数据。

They they want this type of data.

他们已经在通过这种方式变现了。

They're already monetizing it.

这让它们能够理解、训练，甚至进一步实现商业化。

This allows them to understand train it, understand it, monetize it even further.

这是一种非常独特的情况，显然大家都想从中获利，我觉得未来这类场景会越来越多。

And it's a unique situation, right, that obviously people want to take advantage of, and I think we're gonna see more and more of that in such situations.

而且不管怎么说，拥有现有数据的客户也都希望能把这些数据利用起来。

And regardless of that, customers who have existing data want to be able to unlock that existing data as well.

你们在 Eon 都在开发哪些工具，来帮助人们把数据应用到 AI 项目中呢？

What what sort of tooling are you all building at Eon to allow people to make use of their data for AI applications?

你们自己是怎么考虑这个问题的？或者你们的客户都在要求什么样的工具？

Like, how are you thinking about this problem yourselves or what what sort of tools are your customers asking for?

那我们回到问题本身来看。

So let's let's go back from the the the problem statement.

为什么数据处理会有这么多工具？

Why there are so many tools for data and processing?

为什么你需要新工具？

Why why why do you need new tools?

这么多年不是已经有很多优秀的公司了吗？大家都明白数据很重要。

Isn't So already so many great companies been throughout the years, and everyone understands data is important.

这么说吧，以前每个数据团队自己找数据，决定做什么项目，然后拿数据去做事情。

So to put it this way, back in the days, every data team found their own data, decide what project do they have, and get data, do something with that.

非常战术性，他们用一些很棒的公司，比如Fifron、DBT、Monte Carlo，所有现有的数据工具来完成他们的任务。

Very tactical, they were using some great companies like Fifron, DBT, Monte Carlo, all the data tools that exist in order to fulfill their tasks.

而有些数据，他们甚至不知道存在，就被锁住了。

And for some of the data, they didn't even know existed, locked.

为什么会被锁住呢？

It Why was it locked?

因为同一家公司里有很多不同的业务部门负责人。

Because there are multiple business unit owners across the same company.

比如说，你是一家公司的数据团队负责人，你在旧金山，或者我们现在在纽约，我们俩是不同的业务部门负责人。

And let's say you're a data team leader in some company and you are based in San Francisco, or we're now here in New York, and both of us are different business unit leaders.

现在有个叫AI的东西，连老板都在玩ChatGPT。

And now there's this thing called AI, and even the boss is playing with CFGPT.

所以，CEO、董事会和股东们都明白AI是真实存在的。

So, the CEO and board and the shareholders, they all understand that AI is real.

所以他们来找你，跟你说，Elad，我们公司有很多数据。

So, they're coming to you and they tell you, Elad, we have a lot of data in the organization.

我们现在意识到数据是新的石油。

We now realize data is new oil.

我们可以用今天的新工具来激活它。

We can actually activate it with the new tools that we have today.

以前我们做不到。

We couldn't before.

用这些数据做点有用的事，用AI来实现，因为这对我们有价值，而且很酷。

Do something with the data, make it useful, and use AI for that because it's valuable for us and because it's cool.

你能做什么呢？

What can you do?

所以你说，太好了，我以前做过类似的事，我只需要把硅谷每天冒出来的那些新玩意儿都利用起来。

So you say, great, I've done this thing before, I just need to bring to I know all of those new cool things that are coming out every day in Silicon Valley, I can just leverage them.

问题是数据在哪里？

The problem is where's the data?

然后你来找我们，我们有各个业务部门的负责人，你可能认识我们，也可能不认识，但假设你找到了我，我是一个业务部门的负责人，我手里可能有数据，你设法说服我让我给你访问我的数据。

And so, you're coming to us and we have business unit leaders, if you even know us, maybe you don't, but let's say that you somehow get to me, I'm a leader of a business unit, I have data probably, and somehow you convince me to give me access to my data.

但我自己都不知道我有什么数据。

Now, I don't know what data do I have.

我手下有很多人。

I have a lot of people working for me.

过去二十年里，他们的数据分散在多个系统里，有些系统甚至没人真正清楚里面存了什么。

They have data in multiple systems for the last twenty years, some of them systems that no one really understands where.

这些系统里有生产数据，还包含敏感信息。

They contain production data, they possess sensitive information.

你知道的，总有一台服务器，没人知道它是干什么的，但它一直连着电，不管是虚拟的还是物理的，大家都怕关掉它，因为不知道里面有什么。

You know, there's always this server that no one knows what it's doing but it's connected to the power, whether virtually or physically, that everyone's afraid to turn off because we don't know what's in there.

所以我们当时就是这种情况。

So we had all that.

假设我碰巧知道里面有什么。

And let's say that somehow I know what's in there.

现在我得找工程师来，可能还要在安全、合规和生产正常运行时间上做些妥协，才能把数据提取出来，交给你，而且存储方式效率很低，这非常困难。

Now I need to bring engineers and to compromise maybe security and compliance and production uptime and to extract the data, just to give it to you, and store it in a very inefficient manner, it's very hard.

我们意识到这样做有问题，因为我们的动机不一样。

We understood that there's a problem with how this works because we have different incentives.

你被安排去做这件事。

You were tasked with doing that.

我的任务是确保我的系统正常运行，确保数据完整，没有数据泄露。

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.

我不会不小心把CEO的工资放在我的数据里，然后被你们用来训练或后续训练。

I don't accidentally have the salary of the CEO inside my data and it's actually going to used for training or post training by you.

所以我们在Eon用非常不同的方式解决这个问题。

So we at Eon solve it in a very different way.

我们可以帮助你，不是帮我，而是帮你，数据团队负责人，以非常简单的方式找到组织中的所有数据，理解它是什么，分类，映射，理解上下文层，语义层，然后能够持续地从我这里获取所有相关数据，而不影响生产，不影响安全性和合规性，我们实际上会保留所有数据。

We can help you, not me, you, the data team leader, find all the data that's in the organization in a very simple way, understand what it is, classify it, map it, understand context layer on top of that, a semantic layer, and then be able to continuously bring all the data from me that is relevant without compromising production, without compromising security, compliance, we're actually keeping all it.

因为数据已经分类，我知道我不会不小心把你不应该拥有的敏感信息分享给你，最终这些信息会进入你的数据。

And because data is classified, I know that I'm not accidentally going to share with you sensitive information that you shouldn't have eventually in your data.

我们可以以非常经济高效且性能良好的方式做到这一点。

We could do it in a very cost efficient and performant way.

所以你可以从任何地方获取数据，把它带给你，并真正交付。

So you can actually do it from all over the place, bring it to you and actually deliver.

听起来你在解决三四个问题。

So it sounds like there's three or four things that you're solving for.

第一，你在帮人们整合大量的历史数据和当前数据。

One is you're aggregating lots of historical and current data for people.

第二，你能够对个人身份信息或其他他们不想分享的字段进行脱敏处理，或者在上面设置权限。

Number two is you're able to then mask personally identified information or other fields that they don't want necessarily shared or set permissions on top of that.

第三，听起来所有这些都可以暴露给AI模型，供它们使用或应用，而且

And then third is it sounds like all this can then be exposed to the AI models for sort of their uses or applications and

对。

Yeah.

而且我要补充的关键点是，客户已经拥有这些数据了。

And and the key and the key point I'll add is that, customers already have this data.

这有点讽刺。

That's kinda the ironic thing.

现在的客户确实已经拥有这些数据。

Customers today already have this data.

这些数据以不同的形式保存在他们的环境中，但都是被锁定的。

It's kept in their environment in, different forms, but it's, locked.

这些数据是访问不了的。

It's not accessible.

而且通常成本非常高。

And usually it's very, very expensive.

对吧？

Right?

所以我们能利用客户已有的数据，将其转换成这种新的数据基础格式，存储效率更高，同时提供映射、分类、访问控制，并连接到AI工作流中。

So we're able to take what customers already have, convert it into this new data foundation format that's much more it's stored much more efficiently and provide the mapping classification access control, and connect it into the AI workflows.

明白。

Mhmm.

你怎么看待安全问题？

How do you think about security?

最近有很多关于实验室的新闻，他们会用沙盒之类的工具来测试代理，做各种事情。

So there's been a lot of news recently about the labs where they'll have agents, like, sandboxes and do all sorts of things.

你知道，除了代理能力本身，可能还有更广泛的原因在推动这些事情发生。

You know, there may be broader things afoot in terms of why that's happening beyond just the agent capabilities.

比如，谁知道这些东西是怎么搭建或配置的？有时候也不太确定大家到底在多大程度上在推动这些技术。

Like, who knows how these things are set up or configured or, you know, sometimes it's a little bit uncertain whether, you know, there's there's that much how people are pushing these things.

你知道的，目前关于AI安全的话题讨论很多。

You know, fundamentally, there's a lot of discussion of, like, AI security.

在企业级技术栈的背景下，你觉得该怎么看待这个问题？企业该做和不该做的事有哪些？CISA又该如何思考这一切？

How do you think about that in the context of the enterprise stack, what people should do or not do, how CISA should be thinking about all this?

对。

Yeah.

所以到目前为止，大家担心的主要是人为威胁，对吧？

So up until now, the concerns came from human threats, right?

这种情况其实并不新鲜。客户经常来找我们说，嘿，我们遭遇了勒索软件攻击。

So this is not new, where customers would come to us and say, hey, we were exposed by this ransomware attack.

我们在AWS工作期间，就遇到过一家受勒索软件影响非常大的客户。

So during our time at AWS, it was a very large customer that was impacted by ransomware.

当时我们认为他们用了我们的技术——也就是我们负责管理的那套灾难恢复服务——已经得到了全面保护。

We we thought that they were completely protected using our technology, the disaster recovery service that we managed there.

不幸的是，我们后来发现客户以为自己受到了保护。

And we learned unfortunately that the customer thought that they were protected.

其实他们并没有被保护，因为他们没有正确地对资源进行映射、分类和标记，所以实际上没有防护。

They weren't protected because they didn't map and classify and tag their resources properly, it wasn't protected.

结果整个环境有60%被勒索软件攻击了。

And so 60% of the environment was was exposed by ransomware.

这也是我们决定推出Eon的原因之一，就是为了解决勒索软件这类人为威胁带来的痛点。

And that's one of the reasons why we decided to launch Eon and solve that pain point around human threats such as ransomware.

所以我们要能检测到这种情况，比如发现异常的写入模式、熵值变化等，然后进行防护，同时让客户能够快速、精细地恢复数据。

So being able to detect when that happens, look for irregular write patterns and entropy changes and things like that, protect against it, and then also allow customers to recover in a granular fashion and very quickly.

现在更严重的是，同样的威胁来自非人类行为者，也就是AI代理，它们实际上拥有对环境的合法访问权限和合法权限，可以访问某些数据库，然后突然之间，而且现在发生得非常快。

What we're seeing now on steroids is that the same type of threat is coming from non human actors, from AI agents that essentially have legitimate access to the environment with legitimate permissions into such and such databases and all of a sudden and this now happens very rapidly.

一张表突然就被删除了。

A table is all of a sudden dropped.

幸运的是，对我们来说，检测、防护和恢复的方法非常相似，但发生速度极快。

Fortunately, for us, it's a very similar methodology in terms of detecting that and protecting against that and allowing to recover, But the velocity of that happening is extreme.

我注意到，大概六个月前，没人愿意跟我讨论这个。

Something I noted is that, like six months ago, no one would even discuss it with me.

但几个月前，几乎我遇到的每个人，每个公司领导，都跟我说他们要么害怕这种事发生在自己身上，要么就是跟我说话的那个人亲身经历过，这太疯狂了。

But a few months ago, pretty much every person I meet, every leader in a company tells me either they are afraid of that happening to them or it personally happened to that person who was speaking with me, which is crazy.

到处都能看到这种情况。

You see it all over the place.

你能看到真正的恐惧，我不再能决定我的数据上到底在跑什么。

You see real fear from I no longer decide what's really running on my data.

我不再理解了。

I don't know longer understand.

我需要同时准备应对外部威胁，因为你知道，所有新的恶意软件让攻击者更容易来攻击我，但也要应对内部威胁，来自我实际上批准在我的环境中运行的代理。

I need to be prepared for both external threats because, you know, all the new malgars make it much easier for attackers to come to me and attack me, but also from the inside with agents I actually approved running in my environment.

所以这是一个非常非常棘手的时期。

So it's a very very tricky time.

我们需要假设已经被入侵，无论是不是恶意的，并且需要能够处理它并采取相应行动。

We need to assume breach, whether it's malicious or not, and need to be able to handle it and act accordingly.

今天的情况真的很奇怪。

It's a very weird situation today.

嗯。

Mhmm.

对。

Yeah.

你怎么看待更广泛的企业技术栈和智能体？

How do you think about the broader enterprise stack and agents?

你知道，现在的技术栈主要是围绕人来设计的，让人能提出精细的分析性问题。

So, you know, the current stack really evolved around people or humans asking therapeutic fine analytical questions.

所以我们有数据仓库、仪表盘、ETL管道，还有商业智能工具。

So we have warehouses, we have dashboards, we have the ETL pipelines, we have BI.

而智能体的行为可能会更动态、更不一样。

And agents may behave differently and more dynamically.

它们可能能够处理更大规模的数据集。

They may be able to reason over much larger sets of data.

它们可能能访问SAP的SaaS应用、历史数据以及其他各种资源，然后采取行动。

They may have access to SAP SaaS apps and historical data and a variety of other things and then act.

那么，在智能体时代，你觉得数据存储、访问和交互方式会发生哪些变化？

And so what what do you think changes in terms of how you store, access, interact with data in the context of, like, the agentic world?

世界？

World?

或者你觉得还有哪些方面需要改变？

Or what what else do you think needs to change?

仪表盘会消失吗？

Do dashboards go away?

比如，什么会发生变化？

Like, what shifts?

我其实觉得我们会看到更多的仪表盘，因为这是唯一能让我们搞清楚世界上到底发生了什么的方式。

I actually think we'll see more dashboards because this will be the only way to kinda let us figure out what the hell is going on in the world.

因为首先，智能体已经开始编写世界上运行的大部分代码，所以这是间接的。

Because first, agent has started to write most of the code that's running in the world, so that's indirectly.

而且，智能体还会激活其他智能体，这些智能体又去激活别的智能体，要追踪这些非人类身份几乎成了不可能完成的任务。

But also agents activating other agents would activate other agents and trying to keep track of the non human identity or that it becomes almost an impossible task.

组织内部有这么多参与者，人类很难理解责任链条。

So many actors inside the organization when it's so very hard for a human to understand the chain of responsibility.

这也是你看到网络安全公司激增的部分原因。

And this is a part of what you're seeing in the proliferation of cybersecurity companies.

现在你在非人类身份（NHI）领域能看到多少网络安全公司？

How many cybersecurity companies you see in NHI, in non human identity right now?

数不胜数。

An infinite amount.

这背后是有原因的。

And there's a reason for that.

这已经成了头号或二号问题了。

It became number one, number two problem right now.

除此之外，第二件事就是端点。

In addition to that, second thing is endpoint.

你看端点安全，看起来好像很少，但围绕它有很多优秀的公司，就在几年前，端点安全还是一个完全不同的问题，比如TDR。现在，所有发生的事情，你看人们现在在笔记本电脑上运行代理，这些代理有时连接到其他网络，它们还连接到，我觉得，可能OpenClaw连到你的WhatsApp，但也连到你的内部网络，还连到其他应用程序。

You see endpoint security, which looks like it's seldom, so many great companies around it, and world just a few years ago when endpoint was a completely different problem with TDRs, Now everything that's happening, you see people are running agents today on their laptops and the agents sometimes connected to other networks and they are connected to, I think, maybe on OpenClaw connected to your WhatsApp, but also to your internal network and also to other applications.

你会发现IT副总裁、CIO们很难理解他们应该怎么做。

And you see it's very hard for the VP of ITs, for the CIOs to understand what should they do.

一方面，董事会和CEO在推动他们，说：现在就在我的组织里启用AI，别拦着我。

On the one hand, they are being pushed by the board, by the CEO, Enable AI in my organization now, don't block me.

你不能拦着我。

You can't block me.

另一方面，这太吓人了。

On the other hand, it's so scary.

我的意思是，每个人，别想技术人员了，想想非技术人员，他们用Lovable或者其他什么软件为自己构建东西，把公司数据放进去。

I mean, every person, don't even think of technical people, think of the non technical people building something with, let's say, Lovable or any other software that you have, for themselves, putting company data there.

他们甚至没有意识到安全、合规或者谁会使用这些数据，而他们都在使用所有这些新潮的东西。

They're not even aware of things like security or compliance or who is going to use this data, and they're all using all of those new cool things.

所以也许他们构建的代理在使用其他代理，而他们甚至不是技术人员，根本不明白这意味着什么。

So maybe their agents that they are building are using other agents and they're not technical to even understand what it means.

这样一来，组织内部就出现了一整套不受组织规则约束、也不一定在组织内部运行的角色，但他们却在处理属于组织的敏感数据，这些数据可能会暴露给外界，被不当使用，这就成了一个大问题。

So it creates a complete set of actors inside the organization, not bound by the rules of the organization and not necessarily running within the premises of the organization, but handling sensitive data, which is the property of the organization, be exposed to the world, it could be incorrectly used, and becomes a big problem.

组织里的每个人都能成为构建者，这既是好事也是坏事，不管你是社交媒体经理、财务人员，还是法务或财务部门的人。

It's a good thing and bad thing that everyone inside the organization can become builders, whether you're a social media manager, whether you're a finance person, whether you're in legal or finance.

所以这很了不起，但同时也意味着我们正处在一个非常有趣的时代。

So it's amazing, but it's also we live in very interesting times in that perspective.

你觉得现有的数据基础设施有多少能在这场变革中存活下来？

How much of the existing data infrastructure do you think survives all this?

你知道，过去十年里，人们一直在构建和部署各种ETL、数据工程基础设施。

So, you know, there's all the ETL, data engineering infrastructure that, you know, people have been building and deploying over the last, you know, decade.

这些还能继续用吗？

Did did that stick around?

它们会发生变化吗？

Does that shift?

这一切会被颠覆得有多快？

Does that like, how quickly does all this up end?

所以你看，有一个强大的伴随事件几乎要改变一切，因为现在的管道非常有限，每个人都针对自己的问题建了一套解决方案。

So you see there's a strong companion event to pretty much change everything because the plumbing today is very limited and everyone built a solution to their set of problems.

所以，想想会发生什么。

So, think about what happens.

现在，Gonen录完这个播客后下楼，他真的很想喝咖啡。

Now Gonen goes downstairs after recording this podcast and he really wants coffee.

于是，他去商店买了咖啡，用信用卡付了款。

So, he goes to the store and buys coffee and he puts on his credit card.

现在有一笔交易，这笔交易被记录在某个数据库里。

Now there's a transaction and this is written in some database somewhere.

对，没错。

Okay, right.

所以，现在有人需要做的是，他们提取数据，放到某个地方，就这样。

So someone needs to, today what they're doing, they're extracting the data, putting it somewhere, and that's it.

之后某个时候，另一个人会拿这些数据，用另一种方式处理，也就这样。

Someone else at some point takes this data and processes it in some other way, that's it.

所以这些东西之间没有任何关联。

So there's no connection with all of those stuff.

而且每个人的情况都不一样。

And every person is very different.

他们不了解之前发生了什么。

They don't have the context of what happened before.

之所以以前不需要，原因很简单，因为以前数据量不大，而且你只能按照最初的想法去处理数据，所以不需要把所有数据和组织的背景都搞清楚。

And the reason it wasn't, and the reason is very simple, it wasn't so important before to have all the context of all the data and post organization, because you could only do with the data things you really intended to do to begin with.

所以你在处理数据时，脑子里只有一个明确的目的。

So you have a single purpose in your mind when acting on the data.

现在情况完全不同了。

Today, it's very different.

现在你明白了，如果你能聪明地收集和清理所有数据，并确保高效存储，而且能高效地利用这些数据，那么你就能让团队充分利用他们手里的所有数据。

Today, you understand that you can collect if you are able to smartly collect and clean all your data and make sure you store it in an efficient manner, and if you can activate that efficiently, you can let a team go wild with all the data that they have.

而且他们拥有的数据越多，数据质量越高，数据背景越丰富，处理这些数据的团队就能创造出惊人的成果。

And the more data that they have, and the more high quality data that they have, and the more context all that data that they have, the team handling that can create wonders.

想想那些以前想都不敢想的事情。

Think of things which were unimaginable.

比如说，公司里有一个人掌握着纽约所有爱吃汉堡的人的名单，另一个人则有纽约所有爱吃披萨的人的数据库。

Let's say that there's one person in the organization who has a list of all the people in New York who love burgers and another person in the organization who has a database of all the people in New York who love pizza.

他们不知道其实可以找到一份纽约既爱吃汉堡又爱吃披萨的人的名单，因为他们没有合作过。

They don't know they can find a list of all the people in New York who love burgers on pizza because they didn't work together.

现在，如果你把这些数据用于新的用途，发挥新的能力，你就能创造出奇迹。

Now if you use it for posturing, use it for the new capabilities, you can actually do wonders with that.

你可以开始向数据提出聪明的问题，让你的数据回答你。

You can actually start asking intelligent questions, your data intelligent questions.

你可以开始把这些数据用于你自己的目的，这是以前做不到的。

You can start using that for your own purposes and just something that you couldn't do before.

所以我们看到，公司现在收集的数据量比以前多得多。

So we've seen companies, first they're collecting tons more data than before.

数据摄入量简直惊人，尤其是和以前相比。

The amount of data being ingested is absolutely insane, especially comparing to earlier.

我们和其他公司都在持续看到这些趋势，数据量一直在增长。

We see trends continuously, both us and other companies that we're seeing data.

数据增长得越来越夸张，量太大了，而且很多都是那些新智能体产生的。

You see data is growing out of proportions, so much of it, and so much of it is being generated by those new agents.

所以数据里有很多噪音。

So there's a of noise in the data.

同时也有很多价值和噪音。

There's a lot of value and noise as well.

所以你需要一些工具，既能理解来自多个地方的数据，又能清理噪音，确保这些新产生的数据真的能用。

So you need tools that are able to both understand data from multiple locations, clean the noise and make sure all of this data that's been created is actually usable.

旧工具做不到这一点，虽然有些旧工具很厉害，比如Fivetran和DBT，它们都是很棒的公司，但它们非常专门化，只针对特定用途。

It doesn't apply with the old tools that were very some of them were incredible, Fivetran was an incredible company, DBT and so on and so forth, but very niche, very specific tools for that purpose.

所以，这就创造了一个非常有趣、全新的世界。

So, this creates a very interesting, brave new world.

像Databricks这样的公司，在我看来是地球上最厉害的公司之一，它们看到数据越来越多，但不知道数据在哪里，于是帮你整理数据目录并加以利用，但这只是事后补救。

You've seen companies like Databricks, one of the most incredible companies on the planet in my opinion, looking at I have more and more data coming in, I don't necessarily know where it is and help you catalog the data and make use of that, but it's an after effect.

数据已经有了，现在你需要处理它。

You already have the data, now you need to process that.

但他们一直在自我革新，因为他们明白越来越多的数据是由智能体生成的。

But they are reinventing themselves all the time because they understand that more and more data is being generated by agents.

他们觉得，在我看来，如果你打不过他们，就加入他们。

And they thought the way I see it is if you can beat them, join them.

我们会构建自己的智能体，构建自己的数据库。

We'll build our own agents, we'll build our own databases.

他们想要掌控数据的使用方式和创建方式。

And they want to take charge of how data is being used, how data is being created.

这和三五年前其他人的做法完全不同。

And it's completely different than how any other people use that just three or five years ago.

所以，目标就是真正赋能这种建设者文化和智能体文化，自动帮助他们理解数据内容，自动帮助他们摄取数据，而无需为每个应用手动构建管道，同时还要帮助他们保持对生成数据的控制。

So, the goal is really to enable this culture of builders and the culture of agents with the ability to automatically help them understand what's there, automatically help help them ingest the data without having to build manual pipelines for each and every application that is being built, and then also help them maintain control on top of the the data that's created.

有道理。

Makes sense.

而且你看，你手头的数据现在还有个问题，数据多当然好，但很分散，这确实是个麻烦。

And you see with data that you have, there's another problem right now that lots of data is amazing, but it's scattered, which is a sound of problem.

但接下来你得能访问这些数据。

But then you need to access that.

你得为存储付费，当然还有token的费用。

You need to pay for that, for storage and of course, tokens.

而且现在我们已经不是追求token最大化的时代了，不是想着怎么把每个token都用足，因为成本越来越高了。

And we're not in the time of token maxing anymore, not trying to go to actually getting value for every token that we have because it comes more and more and more expensive.

所以你得非常明智，我不是说不要花几百万，如果需要的话花几百万甚至更多也行，但关键是要从中获得应有的价值。

So, you want to be very wise in I don't want to say not paying millions, paying millions and even more than that if you need to, but get the value that you can from actually doing so.

所以这既很贵，又很赚钱。

So it's very expensive, very lucrative.

我们要尽量让它变得尽可能便宜。

Let's make it relatively as least expensive as you can have it.

嗯。

Mhmm.

所以我想，你们真正经历过的另一件事就是云迁移。

So I guess, know, the other thing that you guys have really lived through is the cloud transition.

在创立Eon之前，你们创办了一家叫CloudEndure的公司，后来被AWS收购了。

So prior to Eon, you started a company called CloudEndure that was acquired by AWS.

在AWS，你们确实看到了从本地部署到云端的迁移，规模巨大，就像之前发生的那种代际性的大变革。

And AWS, you really saw that migration from on prem to the cloud at, like, a huge scale in terms of that that big sort of generational issue that had happened before this.

你会怎么比较这次基础设施变革和现在AI正在发生的事情？

How would you compare this infrastructure change to what's happening with AI right now?

比如，你怎么看待云时代和AI时代，有哪些可以互相借鉴的经验或教训？

Like, what what do view as sort of the cloud era versus AI era, what are takeaways or lessons that you can apply across them?

对。

Yeah.

我觉得，这就像是云迁移的加强版。

I think, again, it's, it's like that but on, but on steroids.

而且，在我们把上一家公司Cloudera卖给AWS之前，我们就已经用我们的产品，通过OEM方式集成到Azure和GCP的控制台里，支持了其他超大规模云平台上的大型企业迁移。

And even before we we sold our last company, Cloudera, to AWS, we, supported similar, large scale enterprise migrations with the other hyperscalers with, with Azure and with GCP where our product was was integrated OEM into the console.

所以那些非常大的企业，迁移成千上万甚至几十万台服务器，然后我们看到这些在云端进一步现代化。

So very large enterprises that were moving thousands, tens of thousands or hundreds of thousands of servers, and then we saw those modernized further in in the cloud.

在我们卖给AWS之后，我们作为应用迁移服务的一部分来做这件事。

And after we sold to AWS, we did that as part of the application migration service.

但基本上就到此为止了，而且这需要大量的工作，大量的努力，无论是技术方面还是人力方面。

But that's kind of where it ended and it required a lot of work, lot of effort, both from a technology side, as well as from the human side.

我们现在在AI和智能体这个疯狂的世界里看到的是，这些转型发生得比以前快得多，客户正在失去控制，以至于这变成了一个阻碍因素，而不是推动因素。

What we're seeing now in this crazy world of AI and agents is that those transformations are happening way faster and customers are losing control to a point where that's becoming an inhibitor, not an enabler.

他们停下来，暂停了，因为他们担心事情会出错，数据会泄露，知识产权会外泄。

They're stopping, they're pausing because they're afraid that things might break, that data might leak, that IP might break out.

所以他们拼命地寻找这种对正在发生的事情的理解和控制。

And so they're looking desperately for this level of understanding of what's happening and control.

所以，这来得太疯狂了，也太快了。

So, it came so insane that and so fast.

其中一个原因是，在我看来，云有点抽象，因为很难解释它到底意味着什么。

And one of the reasons is that cloud, in my opinion, is somewhat abstract, because it's very hard to explain what it means.

云基本上就是别人的电脑，但谁知道它到底是什么。

Cloud is basically just someone else's computer, but who knows what it is.

很难跟我奶奶解释什么是云。

It's hard to explain to my grandmother about the cloud.

AI，大家都懂AI。

AI, everyone understands AI.

大家都经历过那个ChatGPT时刻，当时我们都笑AI能做什么，然后惊叹，天哪，这太不可思议了。

Everyone lived from the chattypiki moment when we all laughed what AI could do, and oh my god, this is incredible.

所以他们被高管层、CEO、董事会、股东推着走。

So they're getting pushed by C levels, by CEO, by the board, by the shareholders.

把AI用起来，不然就落伍了。

Use AI for the business, otherwise irrelevant.

所以你会看到人们这么做，既是为了从AI中获得价值，也是因为对AI的恐惧。

So you see people doing it both for the value that you get from AI, also for the from the fear that you get from AI.

而且你会看到新的趋势，多年来第一次，你会看到公司以一种全新的方式消费软件。

And you see new trends of for a first time in a lot of in many years, you see how companies consume software in a brand new way.

一个例子就是现在前沿部署工程师的情况。

One example is what's happening with forward deployed engineers.

以前这只是一些本地服务或顾问在做的事，没人真正明白那是什么意思，现在大家都在做这个。

It used to be something local services, panelists were doing that, no one really didn't understand what it means, and now everyone's doing that.

现在看起来，你进入一个大型传统企业，他们真的很想采用AI，因为问题在于他们不知道怎么去做。

Now it seems that you come into a large legacy enterprise, they really want to adopt AI because the problem is they don't know how to do it.

他们明白自己的流程非常漫长。

They understand that their processes are very long.

有时候要花一两年甚至更久，但他们现在就需要它。

It sometimes takes a year or two or more, but they need to have it now.

他们唯一能真正部署AI并更快实现AI化的方式，就是让那些懂行的优秀工程师，带着他们在硅谷顶尖初创公司甚至大公司里打造的工具箱，来改造这些组织。

And the only way they can actually get it deployed and become AI much faster is by letting strong engineers who understand what they're doing and coming with the toolbox that they've created in top Silicon Valley startups and sometimes larger companies to come and transform those organizations.

你会看到他们缩短销售周期，公司也因此快速成长。

And you see them shrinking sales cycles and you see companies growing really fast because of that.

你还会看到公司快速采购，尤其是新公司，通过产品驱动增长，快速采用AI基础设施，这实际上有助于构建智能体，因为现在每个人都想构建智能体。

You also see companies buying really fast, especially the new companies buying using product led growth, by a really, really fast AI infrastructure which actually helps to build agents because everyone now wants to build agents.

以前我总觉得，对大多数东西来说，PLG行不通，尤其是开发者工具，因为世界太碎片化了，大家不想动作那么快，等等。

Now in the past I was arguing that for the majority of things, PLG doesn't work, especially for DevTools, because the world is very fragmented, people don't want to move so fast, so forth.

但现在，像Cognition这样的公司变得超级火，它是一家很厉害的公司，先通过PLG走通，我们在Eon也这么用，然后通过FD模式转向银行，用我们的工程师去替换那些你不想做的工程，让你专注于真正想做的事情。

Now, it became super hot looking companies like Cognition, for example, which is an an incredible company that were able to first go through PLG, we use that way in Eon, and then through the FD motion turning, going to banks and then will replace engineering that you don't want to do with our engineers, making you focus with the things that you do want to do.

所以，是全方位发力。

So, leveraging on all fronts.

所以，这变得超级超级有意思。

So, it became super, super, super interesting.

世界变化太大了。

The world is changing so much.

还有一个很有意思的点，就是公司利用AI的方式，他们虽然采用AI很慢，但有些很棒的公司，比如Long Lake，会说，与其你慢慢去搞AI，不如我懂怎么更高效地做。

And one other really interesting way that companies are leveraging AI is they are very slow to adopt AI, but there are really great companies, for example, Long Lake, that say instead of you adopting AI, I know how to do it more efficiently.

如果我能把公司买下来，把它变成一家AI公司，那大家都能赢。

If I can buy the company and transport that into an AI company, we can all win.

我们可以打造一个更高效、利润率更高的平台。

We can create an avatarship, make it higher margin more efficiently.

这对那些公司来说，真的是一个全新的方式，让他们开始使用AI并变得更高效。

And this is a really radical new way for those companies to actually start using AI and become more efficient.

我们把这称为一场革命，但我觉得我们才刚刚开始。

And we speak about this as a revolution, but I think we just started.

大多数公司仍然没有使用AI。

Most companies still don't use AI.

大多数公司还处于这个旅程的起点，他们都明白有些事情正在发生。

Most companies still at the beginning of this journey, they all understand that something is happening.

他们都明白数据很重要。

They all understand that data is important.

他们都明白现有的流程有些枯燥，需要做出改变，但这很吓人，不过你必须去做。

They all understand that their existing processes are somewhat mundane and they need to do something about it, but it's scary, but you have to do it.

所以，这真是令人着迷的景象。

So, fascinating it's thing to see.

现在公司如何消费AI软件、如何转型变得更现代化、以及他们被推动着去做这些事的程度，这一切的演变都非常引人入胜。

It's a fascinating evolution on what's going on right now in how companies consume AI software, how companies transform into being more modern, how much they're being pushed to do that.

而且我觉得最终，我知道这是一段非常疯狂的旅程，但我认为每个人都会变得更好，在我看来，因为这一切，世界会变得更好。

And I think that eventually, I know it's a very wild ride, but I think everyone is gonna go warmth in my opinion is gonna be better because of that.

太棒了。

Amazing.

好的，非常感谢你今天来参加我的节目，Ofir Eonen。

Well, thank you so much for joining me today Ofir Eonen.

我们聊了关于数据和人工智能的非常广泛且有趣的话题。

Very interesting wide ranging conversation on data and AI.

真的很感谢。

Really appreciate it.

谢谢。

Thank you.

我们的荣幸。

Our pleasure.

非常荣幸。

Much.