当数据成为护城河：Eon 创始人谈 AI 时代的企业数据基础设施

本期节目来自 No Priors 频道（《AI、机器学习、科技与创业》），主持 Elad Gil，嘉宾是 Eon 联合创始人兼 CEO Ofir Ehrlich 与总裁 Gonen Stein。

两人有一个核心判断：在算力与模型被高度追捧的今天，真正构成企业护城河的，是那些散落各处、被锁死的历史数据——而 AI 时代正把它们从「积灰的磁带」变成抢手的金矿。Google 收购破产的 Spirit Airlines 的数据，正是这一趋势最生动的注脚。

全文为中英对照，中文在上、英文原文在下，按节目时间戳分为十二个小节。

01｜开场预告：智能体正在成为新的威胁　00:00

到目前为止，担忧主要来自人为威胁。而现在我们看到的，是同类威胁以更极端的形式来自非人类行为者——那些本质上拥有环境合法访问权限和合法权限的代理。

Up until now,  the concerns came from human threats. What we're seeing now on steroids  is that the same type of threat is coming from non-human 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，我们采用的方法论非常相似。

Fortunately for us, it's a very similar methodology in terms of detecting that and protecting against that.

但这种情况发生的速度极快。想想那些非技术人员吧。

But the velocity of that happening is extreme. Think of the non-technical people.

他们甚至没有意识到安全、合规或谁会使用这些数据等问题。

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

也许他们正在构建的智能体在使用其他智能体，而他们甚至不具备技术背景来理解这意味着什么。

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

这会在组织内部产生一组完全不受组织规则约束、不一定在组织内部运行但却处理敏感数据的参与者。

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.

组织里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 organization can become builders.

02｜嘉宾与本期概览　00:59

我们生活在非常有趣的时代。今天，在 no prior 节目中，我们邀请到了 Eon 的联合创始人 Afair Erlick 和 Gunnen Stein。

We live in very interesting times. Today on no prior we're joined by Afair Erlick and Gunnen Stein the co-founders of  Eon.

Eon 是一个面向 AI 时代的云备份灾难恢复中心服务。

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

[音乐] 在这次讨论中，我们聊了数据 AI、为什么 Google 收购了破产的 Spirit Airlines 的数据，以及真正管理和使用 AI 时代数据基础设施意味着什么。

（音乐） 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.

A feel Gunnan，非常感谢你们今天来参加我的节目。

A feel Gunnan, thank you so much for joining me under prior today.

很高兴见到你。当然，谢谢邀请我们。是的。

It's great to see you. Absolutely. Thanks for having us. Yeah.

03｜Eon 是做什么的　01:27

所以，你们在 Eon 做的一件事是——或者不如你们先快速介绍一下 Eon 是做什么的，因为我觉得这能为我们如何思考 AI、数据、模型和微调模型奠定背景。

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

我觉得有一整套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 data sets.

那么，也许我们可以从你们的工作开始聊起，然后我想我们会探讨一下世界相对于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 kind of walk  through like how the world is shifting relative to to the enterprise  data stack.

是的，当然。那么从高层来看，我们做的事情是创建了一个运行在云上的新数据基础，并提供多种能力，让客户能够首先在多个超大规模云环境中映射和分类他们的数据，识别他们拥有什么、在哪里拥有、哪些敏感哪些不敏感等等。

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

然后我们提供了一种能力，可以轻松地将来自所有这些不同来源的数据——结构化和非结构化数据——摄取到这个数据基础中。

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

而数据基础层则提供了一种非常经济高效的方式，既能维护数据以进行保护和恢复，又能理解数据的含义。

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.

因此，它能让客户非常轻松地访问、查询和搜索这些数据，并在从多种来源摄取的数据之上应用他们的 AI 模型和 LLM。

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 uh ingested  from a variety of sources.

04｜数据即护城河　02:41

是的。我的感觉是，你们的起点其实是作为备份和数据恢复及保护服务，而我觉得在这个过程中你们意识到，如果从备份的角度拥有所有这些数据，并且拥有客户的全部历史记录，就可以开始将其用于有趣的应用领域。

Yeah. And my  sense is I mean your starting point was really as sort of backup and  data recovery and protection service and I think along the way you kind  of 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.

你们看到客户在哪些方向上利用这些数据，或者说你们所代表的完整数据历史呢？

Um what are  what are some of those directions where you're seeing customers take  this these 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，我说过我是个疯狂的人，在 AI 的世界里创办一家非 AI 公司，而 AI 的顺风车变得极其疯狂，并确保数据成为组织拥有的最重要资产。你可以想想，模型、算力，一切都是相对短暂的，几乎零切换成本，而这些是基础设施的重要部分。

So as you  mentioned uh when we started I said I'm this crazy person starting a  nonAI company in an AI world and the AI tailwind became absolutely  insane and made sure that uh data becomes the most important thing that  an organ organization have when you can think about it um models h  compute everything is relatively ephemeral almost zero switching cost  and those are infra important part of the

行业的基础设施。但如果你是一家公司，无论你是连锁酒店还是食品科技公司，都没关系。

infrastructure  for uh the industry. But if you are a company, whether you hotel chain  or you are food chain technology company, doesn't matter.

你拥有的最宝贵的东西其实就是你的数据。

The most valuable thing that you have is actually your data.

而且你会看到越来越多的公司发现这一点。你知道，就在两天前，谷歌收购了破产的 H Spirit Airlines 的某些资产。

And you see  more and more companies finding this out. You know, just two days ago,  you saw Google buy a something from the H bankrupt H Spirit Airlines.

他们没有买飞机。他们买的是数据。他们花 1000 万美元买下这些数据，因为他们认为从这个角度来看，这些数据非常重要。

They didn't  buy airplanes. They bought the data. They bought the data for $10  million because they think it's very important in that perspective.

他们用这些数据来训练模型。我觉得还有个传闻是，在那个破产竞标过程中，另一个竞标者其实是 Merkor，对吧。

They're using  that to train models. I think the rumor too is that the other bidder on  that data set was Merkor, right, in terms of the bankruptcy bid process.

这很有意思。AI 领域有多家不同的公司在竞标一家破产航空公司的企业数据集，这确实很吸引人。

And so it's  interesting. You had multiple different companies in the AI world  bidding on a bankrupt airlines enterprise data set, which is  fascinating.

是的，你觉得我们未来会看到更多这样的情况吗？

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

你觉得我们基本上会看到这种破产后的数据收购吗？

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

到目前为止，我们在多个用例中都看到了这一点。这正是它的酷炫之处。

So for we've seen it for multiple use cases. That's what's really cool about it.

而且你看 Mercury，还有其他公司，它们一直在尝试，甚至已经在尝试购买数据了。

And you see Mercury see other companies are continually trying to already trying to buy data.

如果你今天是科技数据公司的 CEO，我可以告诉你，你会经常收到各种问题。

If you're a tech data CEO today I can tell you that you constantly get questions.

你愿意出售你的数据吗？我到处都听到这个说法，看起来这将会成为一个重要的趋势。

Are you willing to sell your data? I hear it all over and it seems that's going to be a signific a significant trend as you go.

我听说有些实验室通过华尔街，试图从对冲基金购买数据，并尝试理解如何绘制和分析公司。

I'm hearing  about, you know, labs going through Wall Street and trying to buy data  from hedge funds and try 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存储在磁带之类的介质上。

So you you see a data that was accured throughout the years by companies which was usually like tapes.

通常就是放在架子上积灰，突然之间它就变得非常重要了。现在公司开始意识到，首先，我今天拥有的、能让我区别于其他任何人的就是我的数据，这些数据是金子，实际上我可以利用它为公司创造更多价值，并在 AI 真正到来时继续发展我的业务。

It was usually  you know a a sitting on a shelf collecting dust and all of a sudden  this becomes very important and you see companies now realize that first  what I have today that differentiates me than anyone else is my data  and this data is gold and actually I can actually leverage it to get  more value for my company and to continue building my business when AI  is actually coming and and and and and

嗯，拉平了竞争环境。看起来每个人都能起步，大公司和小公司基本上拥有相同的竞争舞台，而如今公司唯一的真正优势当然是人，还有他们积累的数据，因为每个人都能用上那些酷炫的新工具。

uh flatten the  playing grounds. It seems that everyone can start even large and small  companies basically have the same the same the same playing field and  the only real advantage that company have today is of course their  people but also the data that they've approved because everyone has  access to all of those cool new tools.

05｜数据是新的石油　09:39

是的，这已经变成了一种护城河，我想，我的意思是，人们长期以来一直在说数据是新的石油，而我对此说法一直有点怀疑。

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

嗯，但我觉得现在的情况是，由于后训练和强化学习，像 Applied Comput 这样的公司以及其他一些公司开始提供这类服务，你可以针对特定数据集微调模型或开源模型，看起来人们正在尝试为自己的用例优化这些东西。

Um but I feel  like now what's happening is because of post- training and reinforcement  learning and you know there's companies like applied comput and others  are starting to provide these sorts of services where you can fine-tune  models or open source models against specific data sets like it seems  like people are trying to optimize these things for their own use cases.

我猜像 Spirit Airline 这种情况，是不是为了构建航空公司应用而提供客户支持？

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

比如，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 going to do with this information?

还是别的东西？我是说内部文档之类的，我有点好奇，这强化学习到底是什么，还是说它就是个客服代理？

Is it  something else? It's the internal documents like I'm just sort of  curious like what what is the the reinfor reinforcement learning or is  it like a customer support agent?

06｜用优质数据训练智能体　06:43

对，但想想看，如果你今天想构建 agent，你不能只在实验室里做。

Yeah. But but think if if if you're a if you're trying to build agents today and trying to you can't just build them a lab.

需要在新数据上训练它们，而很难找到非常好的数据集。

need to train them on on on on new data on on some training data and it's very hard to find very good data sets.

你看，哈佛几天前刚发布了一个法律数据集，但真正好的数据集并不多，那些不像真实合成数据、能真正反映现实世界的数据集很难找到。我觉得 Spirit 航空既可以用作航空公司，也可以作为一个大型企业，一个有很多人工作、层级分明的地方。

You see that  Harvard just released a a a legal data set just a few days ago and but  you don't find too many good data sets that doesn't look like real  synthetic data that can actually be used to really look like the real  world and I think that spirit dines can be used both as an airline  company but also as an large enterprise as a place where lots of people  work a lot of you know the hierarchy middle

管理层、高层管理和员工一起协作，而且你知道，如果你看看其他公开数据集，市面上其实没那么多。说真的，我在和公司交流时问他们有什么样的数据、用什么训练，比如 ANG 数据是公开的，人们实际上把它当作真实数据来用。

management top  management and workers working together and you know if you're looking  at what other h public data sets do you have out there there aren't a  lot of those there's the the seriously I'm speaking with companies  asking what kind of data do you have what you train on real fine stuff  for example the ang data is out there in public and people are actually  using that as real data from a

公司是如何运作的，原因在于很难找到能帮助你在现实世界中工作的数据。

company how a  company works like and the reason is it's so very hard to find data that  will help you to work like in the real world.

每当你看到有人在构建 agent 或新应用时，其实大多数都不太管用。

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

你必须走出去，真正与真实世界的公司互动，才能构建出有意义的东西。

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，当你去见客户时，你可以这么做。所以，你可以购买数据并在内部训练。

Now, you can do that when you go to customers. So, you can, you know, 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在最初的互动中就与客户打交道。

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

所以我认为你会看到越来越多这样的情况，既通过创造新的合成数据，以新的创新方式生成合成数据，再加上获取现有数据——无论是真实数据，还是某种可能包含敏感信息（如 PII、财务信息等）的数据——并真正在此基础上构建现实世界的东西。

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

是啊，Google 显然在旅游领域已经有一段时间了，对吧？

Yeah, and Google obviously is uh it's not it's not they're in this travel uh space for a while, right?

他们想要这类数据，而且已经在用它变现了。

They they want this type of data. Uh they're already monetizing it.

这让他们能进一步理解、训练、利用这些数据，甚至更深入地变现，这是个独特的情况，对吧？

This allows them to uh understand, train it, understand it, monetize it even further and it's a unique situation, right?

这显然是人人都想利用的，而且我认为在这种情形下我们会看到越来越多这样的情况。不管怎样，那些拥有现有数据的客户，也希望能够解锁这些数据。

That uh  obviously people want to take advantage of and I think we're going to  see more and more of that in such situations and regardless of that  customers uh who have existing data want to be able to unlock that  existing data as well.

07｜重新构想数据基础设施：解锁被锁死的企业数据　22:11

你们在 Eon 都在构建什么样的工具，让人们能够利用他们的数据来开发 AI 应用？

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

你们自己是怎么思考这个问题的，或者说你们的客户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?

所以我们回到问题本身，以及为什么会有这么多数据处理工具。

So let's let's go back from the the the the problem statement and why there are so many tools for data and processing.

为什么你需要新工具？这么多年不是已经有很多优秀的公司了吗，而且大家都明白数据很重要？

Why why do you  need new tools? Isn't it sold already so many great companies  throughout the years and everyone understand data is important?

所以这么说吧，以前每个数据团队都能找到自己的数据，决定要做什么项目，然后拿到数据去做点事，非常战术性的。他们会用一些很棒的公司，比如  FAR、DBT、Monte  Carlo，以及所有现有的数据工具来完成他们的任务。而有些数据他们甚至不知道存在，这些数据是被锁定的，为什么会被锁定呢？

So to put it  this way um back in the days every data team could find their own data  decide what project do they have and you know get data do something with  that very tactical they were using I don't know some great companies  FAR DBT Monte Carlo all the data tools that exist you know in order to  fulfill their tasks and for some of the data they didn't even know exist  it was this was locked why was it locked

因为同一家公司里可能有多个业务单元负责人，假设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单元负责人，现在有了 AI 这个东西，连老板都在玩 ChatGPT，所以 CEO 和

because there  are multiple business unit owners across the same company and let's say  you're a data team leader in in some company and you are based in San  Francisco or we're now here uh in New York and both （清嗓） of us different  business unit leaders and now there's this thing called AI and even the  boss is playing with chef GPT so the CEO and the sh and

董事会和股东们明白，AI 是真实存在的。

the board and the shareholders They understand that AI is real.

所以他们来找你，告诉你，我们组织里有很多数据。

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

我们现在意识到数据是新的石油。借助我们今天拥有的新工具，我们确实可以激活它。

We now realize data is new oil. We can actually activate it with the new tools that we have today.

我们以前无法利用这些数据做点什么，让它变得有用，并用 AI 来实现，因为这对我们有价值，而且很酷。

We couldn't  before 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？所以你说，太好了，我以前做过这件事。

What can you do? So you say great I've done this thing before.

我只需要把硅谷每天冒出来的那些新酷玩意儿都带过来。

I just need to bring to I I know all of those new cool things that coming out every day in Silicon Valley.

我可以直接利用它们。问题在于数据在哪里？所以你们来找我们，而我们是业务部门的负责人。

I can just leverage them. The problem is where's the data? And so you come to us and we are business unit leaders.

如果你认识我们，也许你不认识，但假设你不知怎么找到了我。

If you even know us, maybe you don't, but let's say that you find somehow got to me.

我是一个业务部门的负责人，我有数据，可能吧，不知怎么的你说服我让我访问我的数据。

I'm a a leader of a business unit. I have a a data probably and somehow you convince me to give me access to my data.

现在我不知道我有什么数据。我有很多人为我工作。

Now I don't know what data do I have. I have a lot of people working for me.

过去 20 年，他们的数据分散在多个系统里。其中有些系统，没人真正清楚里面存了什么生产数据，因为涉及敏感信息，你懂的，总会有那么一台服务器，没人知道它在跑什么，但不管是虚拟还是物理的，都连着电，谁都不敢关掉它，因为不知道里面装了什么。

They have data  in multiple systems for the last 20 years. Some of them system that no  one really understands where they contains production data because  sensitive information you know there's always this server that no one  knows what it's doing but connected to the to the power whether virtual  or physically that everyone's afraid to turn off because we don't know  what's in there.

所以我们拥有所有这些，假设我不知怎么知道里面有什么。

So we had we had all all of all of that and let's say that somehow I know what's in there.

现在我需要带上工程师，可能还要在安全、合规以及生产运行时间上做出妥协，仅仅是为了把数据提取出来交给你，并以一种非常低效的方式存储它。

Now I need to  bring engineers and and compromise maybe security and compliance and and  uh production up time and to extract the data just to give it to you  and store it in a very inefficient manner.

这非常难。我们意识到这种方式存在问题，因为我们的激励不同。

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

你的任务是做那个。我的任务是确保我的系统正常运行，并确保数据完好无损。

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.

数据不会自己跑掉。我不会不小心把 CEO 的薪资放在我的数据里，然后它实际上会被你用来做训练或后训练。

No data is  running away. I don't accidentally have the salary of the CEO inside my  data and it's actually going to be train to be used for training or post  training by you.

所以我们用非常不同的方式解决这个问题。我们可以帮你——不是帮你，而是帮数据团队负责人——以非常简单的方式找到组织中的所有数据，理解它、分类它、映射它，理解上下文，在此基础上构建语义层，然后能够持续引入所有相关数据，而不影响生产，也不

So we aton  solve it in a very different way. We can help you not me you the data  team leader find all the data that's in organization in a very simple  way understand what it is classify it map it understand context layer on  top of that as build a semantic layer and then be able to continuously  bring all the data from me that is relevant without compromising  production without

在妥协安全合规的同时，我们实际上保留了审计，因为数据已分类。

compromising security compliance we're actually keeping audit and because data classify.

我知道我不会不小心把你不该最终拥有的敏感数据分享给你。

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

我们可以用非常经济高效的方式做到这一点。所以你可以从任何地方获取数据，把它带过来，然后实际使用。

We can do it  in a very costefficient and performant way. So you can actually do it  from all over the place, bring it to you and actually use.

听起来你们在解决三四个问题。

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

一个是你在为用户聚合大量的历史和当前数据。

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

第二点是，你能够屏蔽个人身份信息或其他他们不一定希望共享的字段，或者在此基础上设置权限。

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.

然后第三点是，听起来所有这些都可以被暴露给 AI 模型，供它们使用或应用于各种场景。

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

而且，关键的一点是，客户已经拥有这些数据了。

And yeah and the key and the key point to that is that customers already have this data.

这就是有点讽刺的地方。今天的客户其实已经拥有这些数据了。

That's kind of the ironic thing. Customers today already have this data.

它以不同形式保存在他们的环境中，但它是被锁定的。

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

它不可访问，而且通常非常非常昂贵，对吧？

It's not accessible and usually it's very very expensive. Right?

所以我们能够利用客户已有的数据，将其转换为这种新的数据基础格式，这种格式存储效率更高，并提供映射、分类，

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

访问控制，嗯，然后把它接入 AI 工作流。你怎么看待安全问题？

access control, um, and connect it into the AI workflows. How do you think about security?

所以最近有很多关于实验室的新闻，说它们会有像逃出沙箱这样的代理，做各种事情，而且你知道，除了代理能力之外，可能还有更广泛的原因导致这种情况，比如谁知道这些东西是怎么设置或配置的，或者你知道，有时候不太确定是否真的存在

So there's  been a lot of news recently about the labs where they'll have agents  like it's escape sandboxes and do all sorts of things and you know there  may be broader things of foot in terms of why that's happening beyond  just the agent capabilities like who knows how these things are set up  or configured or you know um sometimes it's a little bit uncertain  whether you know there's

人们处理这些事情的方式有很多不同，但你知道，从根本上讲，关于 AI 安全有很多讨论。在企业技术栈的背景下，你怎么看待这个问题？人们应该做什么，或者不应该做什么？

there's that  much uh uh how people are approaching these things but you know  fundamentally there's a lot of discussion of like AI security um how do  you think about that in the context of the enterprise stack what people  should do or not do,

08｜自主安全威胁：来自智能体的攻击　15:00

CISO 应该如何思考这一切。是的。到目前为止，担忧主要来自人为威胁，对吧？

how CISO should be thinking about all this. Yeah. So, up until now, the concerns came from uh human threats, right?

所以，这并不是什么新鲜事，客户会来找我们说：“嘿，我们遭到了勒索软件攻击，数据被暴露了。”

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

所以，在 AWS 工作期间，有一个非常大的客户受到了勒索软件的攻击。

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

我们当时以为，通过我们管理的灾难恢复服务和技术，他们已经得到了完全的保护。

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

不幸的是，我们了解到客户以为他们受到了保护。

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

它们之所以没受到保护，是因为没有妥善地映射、分类和标记自己的资源。

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

所以它没有受到保护。结果环境中 60% 的部分被勒索软件暴露了。

So it wasn't protected. And so 60% of the environment was uh was exposed by ransomware.

这也是我们决定推出 Eon 并解决勒索软件等人为威胁这一痛点的原因之一。

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

所以能够检测到这种情况发生，寻找不规则的正确模式和熵变化等等，对此进行防护，然后还能让客户以细粒度的方式快速恢复。

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

我们现在看到的情况变本加厉了，同样的威胁正来自非人类行为者，也就是 AI 智能体，它们本质上拥有合法访问权限，可以合法地进入某些数据库环境。

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

然后突然间——现在这种情况发生得非常快——一张表突然就被删掉了。

and all of a sudden and this now happens very rapidly uh a table is all of a sudden dropped.

幸运的是，对我们来说，检测、防护和恢复的方法论非常相似，但这种情况发生的速度极快。

Uh 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.

09｜智能体如何改变企业技术栈　18:15

是啊，我注意到的是，大概 6 个月前，根本没人愿意跟我讨论这个。

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

但几个月前，几乎我遇到的每个人，每家公司的领导，都告诉我他们要么害怕这种事发生在自己身上，要么这种事就亲身发生在了和我说话的那个人身上。

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 per to to the the that person who speaking  with me,

这太疯狂了。到处都能看到这种情况。你似乎真的害怕，我不再决定我的数据上真正运行的是什么。

which is crazy. You see it all over the place. You seem real fear from e I no longer decide what really running on my data.

我不再理解了。我需要同时防范外部威胁，因为你知道，所有新的劫匪让攻击者更容易来找我并攻击我。

I don't know  longer understand. I need to be prepared for both external threats  because you know all the new muggers make it much easier for attackers  to h come to me and and and attack me.

但也要防范内部，因为我实际批准了代理在我的环境中运行。

but also from the inside with agents I actually approved running in my environment.

所以这是一个非常棘手的时期。我们需要假设存在入侵，无论是否是恶意的，并且需要能够处理它并采取相应行动。

So it's a very  very tricky time. We need to assume breach whether it's malicious or  not and need to be able to handle it and act accordingly.

今天的情况非常奇怪。是的。你如何看待更广泛的企业技术栈和智能体？

It's very weird situation today. Yeah. How do you think about the broader um enterprise stack and agents?

所以你知道，当前的这套技术栈确实是围绕着人来演进的，也就是人类提出非常明确的、定义好的分析性问题。

So you know the current stack really evolved around people or humans asking very defined analytical questions.

所以我们有数据仓库、仪表盘、ETL 管道、BI，而智能体可能会以不同且更动态的方式运作。

So we have  warehouses, we have dashboards, we have the ETL pipelines, we have BI  and agents may have may behave differently and more dynamically.

它们可能能够对更大规模的数据集进行推理。

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

它们可能能够访问  SAP、SAS  应用和历史数据以及其他各种内容，然后采取行动。那么，在智能体时代，你认为在存储、访问和交互数据方面会发生哪些变化？或者你认为还需要改变什么？仪表盘会消失吗？会有什么转变？我实际上认为我们会看到更多的仪表盘，因为这将是唯一一种

They may have  access to SAP SAS apps and historical data and a variety of other things  and then act 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  or what what else do you think needs changed do dashboards go away like  what shifts I actually think we'll see more dashboards because this  will be the only way to kind of

搞清楚他们在世界上搞什么名堂，因为第一个编程智能体已经开始编写运行在墙上的大部分代码了。

figure out  what they have going on in the world because first coding agent has  started to write most of the code that's running in the wall.

所以这间接地，但也是智能体激活其他智能体，再激活其他智能体，试图追踪非人类身份，这几乎成了不可能完成的任务。

So that's  indirectly 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 impossible task.

组织内部有这么多行动者，人类很难理解责任链，这就是你看到网络安全公司激增的部分原因。

So many actors  inside the organization when it's so very hard for a human to  understand the the the chain of responsibility and this is a part of  what you're seeing in proliferation of of cyber security companies.

现在你在非身份认证的网络安全领域能看到多少公司？数不胜数，这是有原因的，它已经成了头号或二号问题。除此之外，第二件事是端点安全，看起来已经解决了，围绕它也有很多优秀公司，而就在几年前，端点安全还是完全不同的问题。

how many cyber  security companies you see in in NHI in non identity right now an  infinite amount and there's a reason for that it became number one  number two problem right now in addition to that second thing is  endpoint you see endpoint security which looked like it solved them  there so many great companies around it and just you know a few a few  years ago when endpoint was a completely different problems with

现在  TVRS，所有正在发生的事情，你会看到人们今天在笔记本电脑上运行 agent，这些 agent 有时连接到其他网络，它们连接到，嗯，比如  open claw 连接到你的 WhatsApp，但也连接到你的内部网络以及其他应用程序，你会发现这对……来说非常困难

TVRS Now uh  everything that's happening you see people are running agents today on  on the laptops and the agents sometimes connected to other networks and  it connect they are connected to uh think on open claw connected to h to  to your WhatsApp but also to your internal network and also to other  applications and you see it's very hard for the for

IT 副总裁是给 CIO 们看的，让他们明白自己该怎么做。一方面，董事会和 CEO 在推着他们，说现在就要在我的组织里启用 AI，别拦着我，你也拦不住我。另一方面，这真的很吓人。我是说，每个人——先别想技术人员，想想非技术人员——都能用某个工具做出东西来。

the VP of IT  is for the CIOS to understand what should they do on the one hand they  want to they are being pushed pushed by the board by the CEO enable AI  in my organization now don't block me you can't block me on the other  hand it's so scary I mean every person don't even think of technical  people think of the non-technical people building something with you  know one

比如说，一个可爱的或者任何其他软件，他们自己把公司数据放进去。

let's say a lovable or or any other software that you for themselves putting company data there.

他们甚至没有意识到安全、合规或者谁会用这些数据，而且他们都在用所有这些新酷的东西。

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

所以也许他们正在构建的智能体在使用其他智能体，而他们甚至不具备技术背景来理解这意味着什么。

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

所以，这会在组织内部创造出一整套不受组织规则约束、也不一定在组织场所内运行的行动者，但他们处理着属于组织的敏感数据，这些数据可能会暴露给外界，可能不正确，也可能被错误使用，从而成为一个大问题。

So it creates a  complete set of actors inside an 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 could be exposed to the world it could be incorrect could  be incorrectly used and becomes a a big problem.

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 organization can become builders.

无论你是社交媒体经理、财务人员，还是法务或财务部门的人。

Whether you're a social media manager, whether you're a a finance person, whether you're in in in in legal or finance.

嗯，确实很惊人，但我们也生活在非常有趣的时代。

Mhm. 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 um 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的 ETL 数据工程基础设施，对吧。

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

这个想法保留下来了吗？它会转变吗？会改变吗？比如这一切多久会被颠覆？

Did that stick around? Does that shift? Does that change? Like how quickly does all this up end?

所以你看，现在有一个强有力的推动因素，促使几乎要改变一切，因为今天所谓的管道系统非常有限，每个人都针对自己的问题集构建了解决方案。

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

那么想想现在会发生什么。Gonen 录完这个播客后下楼，他真的很想喝咖啡。

So think about what happens now. Gonen goes downstairs after recording this podcast and he really wants coffee.

于是他去商店买了咖啡，并用信用卡支付。

So go to the store and 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.

好的，没错。所以现在有人在做的是，他们把数据提取出来放到某个地方，就这样。

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

某个时候，其他人会以某种其他方式获取这些数据并进行处理，就这样。

someone else at some point takes this data and process in some other way and that's it.

所以这些东西之间没有关联，而且每个人都非常不同。

So there's no connection with all of those stuff and every person is very different.

他们不了解之前发生了什么，也不知道为什么没有，原因很简单。

They don't have the context of what happened before and the reason it wasn't and reason is very simple.

以前拥有所有上下文和数据结构并不那么重要，因为你只能对数据做你一开始就真正打算做的事情。

It wasn't so  important before to have all the context all the data 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，脑子里只有一个目的。

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

嗯，嗯。今天情况大不相同了。如今你明白，如果你能聪明地收集并清理所有数据，确保以高效的方式存储，并且能高效地激活这些数据，

Mhm. Mhm.  Today it's very different. Today you understand that you can collect if  you're able to smartly collect and clean all your data and make sure you  store in efficient manner and if you can activate that efficiently,

你可以让一个团队尽情利用他们拥有的所有数据。

you can let a team go wild with all the data that they have.

他们拥有的数据越多，数据质量越高，对数据的上下文了解越深，处理这些数据的团队就能创造出奇迹。

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

然后想想那些以前无法想象的事情。比如说，组织里有一个人掌握着纽约所有喜欢汉堡的人的名单，另一个人则拥有一个数据库，记录着纽约所有喜欢披萨的人。

And think of  things which were unimaginable. Let's say that there's one person in  organization who have all the list of all the people in New York who  love burgers and another person in the organization have who has a  database of all the people in New York who love pizza.

他们不知道，他们可以找到一份纽约所有喜欢在披萨上加汉堡的人的名单，因为他们没有合作过。

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

现在，如果你把它用于海报，并利用它的新功能，你实际上可以创造出惊人的效果。

Now if you use it for poster and use it for a for the new capabilities you can actually do wonders with that.

你实际上可以开始向你的数据提出智能问题。

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

你可以开始将它用于自己的目的，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.

所以你会看到公司首先收集的数据比以前多得多。

So you're seeing companies first they're collecting tons more data than before.

数据摄入量绝对惊人，尤其是与早期相比。

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

我们和其他公司都持续看到数据方面的趋势。

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

数据增长已经失控了。数据量巨大，而且这些新 agent 生成了海量数据。

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

所以数据里有很多噪音，也有很多价值，噪音也一样多。

So there's a lot of a lot of noise in the data. there's a lot of value and noise as well.

所以你需要一些工具，既能理解来自多个位置的数据，又能清理噪音，确保这些产生的数据真正可用。

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

嗯，而且这不适用于那些旧工具，其中一些非常出色。

Mhm. And it doesn't apply with the old tools that were very some of them were incredible.

Five 是一家很棒的公司，DBT 等等，但这些都是非常针对性的工具。

Five was an incredible company, DBT and so on so forth, but very specific tools for that purpose.

所以这创造了一个非常有趣的勇敢新世界。你见过像 Databricks 这样的公司，在我看来，它们是地球上最令人难以置信的公司之一，而且我看到越来越多的数据涌入。

Uh so so this  creates a a very interesting brave new world. You've seen companies like  data bricks, you know, one of the most incredible companies on the  planet in my opinion and looking at I have more and more data coming in.

嗯。我不一定知道数据在哪里。我会帮你编目数据并加以利用，但这是事后效应。

Mhm. I don't necessarily know where it is. I'll help you catalog the data and make use of that, but it's an after effect.

你已经有了数据。现在你需要处理它。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，而他们认为，在我看来，如果你打不过他们，就加入他们——我们会构建自己的智能体，我们会构建自己的数据库。他们想要掌控数据的使用方式和创建方式。

You already  have the data. Now you need to process that. H but they are reinventing  themselves all the time because they understand that more and more data  has been generated by agents and they thought the way I see it is if you  can't beat them join them we'll build our own agents we'll build our  own databases we got we they want to take charge of how data is being  used data is being created

这和三五年前任何其他人的用法完全不同。

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

是的。所以目标确实是赋能，赋能这种构建者文化和智能体文化，让它们能够自动理解那里有什么，自动摄取数据，而不必为每个正在构建的应用手动搭建管道，同时还要帮助他们在所创建的数据之上保持控制。

Yeah. So the  goal is really to enable right enable this culture of of builders and  the culture of uh of agents with the ability to automatically help them  understand what's there automatically 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。而且你看，你拥有的每一份数据都伴随着另一个问题——现在数据量庞大固然很好，但它们分散各处，这本身就是一种问题。然后你需要访问这些数据，需要为存储付费，当然还有  token 费用。而且我们不再处于 token 最大化的时代了，而是要努力让每一个 token 都真正产生价值，因为

Makes sense.  And and you see every with every data that you have there's another  problem right now that lots of data is amazing but it's scattered which  is a sort of problem but then you need to access that you need to pay  for that for storage and of course tokens and we're not in the time of  token maxing anymore trying to go to actually getting value for every  token that we have because

成本会变得越来越高，越来越昂贵。

comes more and more and more imper more and more and more expensive.

所以你要非常明智，不是说不要花几百万，甚至更多如果需要的话，但要从中获得实际的价值。

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

所以这非常昂贵，也非常有利可图，让我们尽可能让它相对便宜一些。

So so it's very expensive very lucrative let's make it relatively as least expensive as you can have it.

10｜云时代 vs AI 时代　27:52

所以我想，你们真正经历过的另一件事就是云迁移。

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

在创立 Eon 之前，你们创办了一家名为 Cloud Endor 的公司，后来被 AWS 收购。

So prior to Eon, you started a company called Cloud Endor 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转变一样。

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

你会怎么把这次基础设施的变革和现在 AI 的发展做比较？

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 时代的区别，以及有哪些经验或教训可以跨时代应用？

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

对，我觉得嗯，这就像是那个，但是嗯，但是是加强版。

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

嗯，甚至在我们把上一家公司 CloudEndor 卖给 AWS 之前，我们就已经支持了类似的大规模企业迁移，与其他超大规模云服务商合作，比如 Azure 和 GCP，我们的产品以 OEM 方式集成到了他们的控制台中。

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

嗯，所以非常大的企业，它们迁移了数千、数万甚至数十万台服务器，然后我们看到这些在云中进一步现代化。

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

在我们卖给 AWS 之后，我们把它作为应用迁移服务的一部分来做。

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

但也就到此为止了，这需要大量的工作和努力，无论是技术层面还是人力层面。

Uh but that's  kind of where it ended and it required a lot of work, a lot of effort  both from a technology side as well as from uh 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，这些转变发生得更快，客户正在失去控制，以至于这变成了一个阻碍因素。

What we're  seeing now in the this crazy world of uh of AI and agents is that those  transformations are h happening way faster and uh and customers are  losing control to a point where that's becoming an inhibitor,

对吧？不是推动者。他们在停下、在暂停，因为害怕东西会坏掉、数据会泄露、知识产权会外泄，所以他们拼命想找到这种对正在发生的事情的理解和控制。

right? Not an  enabler. they're stopping, they're pausing because they're afraid that  things might break, that data might leak, that uh IP might uh break out  and uh and so they're looking desperately for this level of of  understanding of what's happening and control.

所以事情变得如此疯狂、如此之快，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。

So it became  so insane that and so fast and one of the reasons is that cloud in my  opinion cloud is somewhat 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呢。

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

很难跟我奶奶解释云 AI 是什么，但人人都懂 AI。

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

每个人，每个人都从 ChatGPT 时刻开始，那时我们都看到了 AI 能做什么。

Everyone, everyone lived from the CHP moment when we all left what AI could do.

天哪，这太不可思议了。所以他们被海平面、被 CEO、被董事会、被股东推动着，用 AI 来经营业务，否则我们就会过时。

Oh my god,  this is incredible. So they getting pushed by by sea levels, by CEO, by  the board, by the shareholders use AI for the business otherwise  otherwise we're relevant.

所以你会看到人们这样做，既是为了从 AI 中获得价值，也是因为对 AI 的恐惧。而且你会看到新的趋势，多年来第一次，公司以一种全新的方式消费软件。前向部署工程师正在发生的事情，以前看起来像是……

So you see  people doing it both for the value that you get from AI but also for the  from the fear that you get from AI and you see new trends of for a  first time in in a lot of in in in in many years you see how companies  consume software in a brand new way what's happening with forward  deployed engineers used to be something look like

Palenteer 服务当时就在做这个。没人真正理解这意味着什么，现在每个人都在做。

services palenteer were doing that. No one really did understand what it means and now everyone's doing that.

11｜AI 正在如何改变企业　30:26

现在看起来，你进入一家大型传统企业，他们真的很想采用 AI，因为他们不得不这样做。

Now it seems that you come into a large legacy enterprise, they really want to adopt AI because they have to.

问题是他们不知道怎么做。他们明白自己的流程非常冗长。

The problem is they don't know how to do it. They understand that their processes are very long.

有些项目需要一年、两年甚至更久，但他们现在就需要。

They some takes a year or or two or more, but they need to have it now.

他们真正能更快部署并实现 AI 化的唯一途径，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，来改造这些组织。

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

你会看到销售周期缩短，公司也因此飞速成长。

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

你还会看到公司购买得特别快，尤其是新公司，通过产品驱动增长（PLG）快速购买 AI 基础设施，这实际上帮助构建了智能体，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 uh buying  using productled growth by in really really fast a AI infrastructure  with which actually h helped to build agents because everyone now want  to build agent.

过去我一直在争论，对于大多数事情来说，PLG  并不适用，尤其是开发工具，因为世界非常碎片化，人们不想那么快行动等等。现在它变得超级热门，比如 Cognition  这样的公司，你知道，这是一家令人难以置信的公司，他们能够首先通过 PLG 方式，我们在 Eon 也是这样用的，然后通过 FD 模式，

Now in the  past I was arguing that for the majority of things PLG doesn't work  especially for dev tools uh because the world is very fragmented people  don't want to move so fast so forth now it became super hot looking  companies like cognition for example which is you know incredible  company that were able to first go through a a PLG we use that that way  in Eon and then through the FD motion telling

去银行，然后用我们的工程师替换掉你不想做的工程，让你专注于你想做的事情。

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.

所以在各个方面都发力，这变得超级超级超级有趣。

So leveraging on all fronts so it became super super super interesting.

世界变化如此之快，而公司利用 AI 的另一个真正有趣的方式是，它们采用 AI 的速度很慢，但有一些非常优秀的公司，比如 Long Lake，它们说，与其让你自己去采用 AI，我更知道如何更高效地做到这一点。

The world is  changing so much and one other really interesting way that companies are  leveraging AI is it's 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.

如果我能收购这家公司并将其转型为一家 AI 公司，我们就能实现共赢。

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

我们可以利用套利机会，获得更高的利润率，提高效率，这对那些公司来说是一种真正革命性的新方式，让它们真正开始使用  AI，变得更高效。我们把这称为一场革命，但我认为我们才刚刚开始。大多数公司仍然不使用  AI，大多数公司仍处于这一旅程的起点。它们都明白有些事情正在发生，它们明白数据是

we can attract  arbitrage make higher margin more efficiently and this is a really  radical new way for those companies to actually start using AI become  more efficient and we speak about this as a revolution but I think we  just started most companies still don't use AI most companies still at  the beginning of this journey they all understand that something is  happening they understand the data is

重要的是，他们明白自己现有的流程有些单调，他们需要对此采取行动，但这很可怕，不过你必须去做。

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

所以，这是一个非常有趣的现象。这是一个引人入胜的演变过程，关于当前公司如何消费 AI 软件，公司如何转型为更现代化的形态，

So, it's a  it's a fascinating thing to see. It's a fasc fascinated evolution on  what's going on right now in how how companies consume AI software, how  conso companies transform into being more modern,

以及我们被如何推动去这样做。我认为最终，我知道这是一段非常疯狂的旅程，但我相信每个人都会走向——在我看来，世界会因此变得更好。

how much we  being pushed to do that. And I think that eventually I know it's a very  wild ride but I think everyone is going to go the world in my opinion is  going to be better because of that.

12｜结语　34:31

太棒了。非常感谢你今天加入我们。这是一场关于数据和 AI 的非常有趣且广泛的对话。

Amazing. Well, thank you so much for joining me today. Very interesting wide ranging conversation on data and AI.

相关链接

• ✓

  节目原视频：https://youtube.com/watch?v=YscDZpVF4CQ

• ✓

  No Priors 频道：https://www.nopriors.ai

• ✓

  Eon 官网：https://www.eon.io

• ✓

  主持人 Elad Gil：https://twitter.com/EladGil

• ✓

  Eon：https://twitter.com/Eon_io_

原节目：No Priors《Rethinking Legacy Data Infrastructure with Eon Co-Founders Ofir Ehrlich and Gonen Stein》