Back to Index

Vibe Coding with Confidence — Itamar Friedman, Qodo


Chapters

0:0 The Evolution of AI in Development
3:8 The Rise of the Command Line Interface (CLI)
3:50 AI Across the Software Development Life Cycle (SDLC)
6:40 The Importance of "Vibe Coding with Confidence"
8:15 The Role of Workflows and Agents
12:21 Qodo's Multi-Agent Approach
13:55 Why the CLI is the Future of AI in Development
20:33 The Future: A "Swarm of Agents"

Transcript

Okay, so my name is Itamar Friedman, CEO and co-founder of Codo, and I'm going to talk about, yeah, Vibe Coding, but more on the confidence part, and why, this is maybe surprising, why CLI might be the new interface leader for that, and we are soft launching a CLI tool right now, five minutes ago, so you can go and install it, it's soft launch, play with it, give us feedback, there is some capacity, and in order to verify who is in the audience, just like if you can tell me if you tried Cloud Code or Codec CLI, okay, and I guess if I ask about AIDEs, you would also do that, right, and the question is why, and I'm going to talk about it a bit, okay.

So the agenda is that, I'm actually going to really zoom out to the Gen 3.0, what do I mean by that, and from the autocomplete to the multi-agent workflow, and workflows, you see it's a theme that's returning here, CLI is something that can support that, and how do we believe that Codo is going to be the real shift in software development, and like why is the CLI is the future, and you know, being able to do vibe coding with confidence, and eventually I'm going to connect, do zoom out again, I'm going to zoom in, zoom out, even with a demo, and connect the dots, and I will say what, explain a bit what is Codo, I didn't do it right now, and we'll do it a bit later, but that's something I'm going to cover shortly.

So AI for developers, and I just think that most people here are developers according to the amount of hands that were raised, so I think like we first got some tools that we liked, AI tools, then we got some tools that we loved, and I'm not that sure that we actually get tools into our hands that completely become a game changer for us.

Okay, it's maybe unpopular opinion, I'm going to elaborate a bit, so in order to do that, I'm going to differentiate developers between noobs and enterprise, it's a bit weird, like it's not necessarily juniors, like there are juniors in enterprise, right, but still, you know, like let's use that for a second, noobs versus enterprise, so I think gen 1.0 was the autocomplete moment, right, we started being able to generate writing our code and getting a few lines ahead of us, we really liked that, if you weren't a developer, like you couldn't really do more, right, so you needed to be a developer already, and then came the gen 2.0, the chat, the agentic chat, and I think like we saw even junior developers generating much more lines of code, maybe the seniors or the enterprise liked it, but also had a lot of problems with it, they need to review much more code, not necessarily high quality code, I hear that a lot, and I'm going to show you like 500 tweets about it, not really, but from some opinion leaders, and I think that the game changer moment is when we will move from chat, agentic chat, to actually being able to, and as a spoiler, command, command line interface, to give commands to our agents as if they're like team members, and they do end-to-end flow, what is an end-to-end flow is something I'm going to talk about a bit, and I think like you will see different players and different parts of this map, like I mean different tools, let's think about it again, so I think like in the an autocomplete moment, we saw first ID plugins, again it's not a game changer, I'm just like going to complete more lines of code for you, we liked it, and then let's rethink how we do coding inside ID, so we saw ID forks, and then what's the third wave, sometimes I hear forking people, I didn't say that, what I think is actually we're going to see is like AI across the STLC, okay, like being able to do end-to-end flows that are specific, closing a task that is specific in the STLC, but more importantly, using AI across the STLC, again, not only in the ID, hence, how do we do that, command line, right, it's one of those tools, a lot of spoilers, so I think when we had the aha moment with the autocomplete, the next aha moment was, you know, all the hype around vibe coding, and I'm going to touch about that, and I think the next thing that I think we're going to hear about is like, hey, I'm really doing 100x, choose a number, 10x, 50x, 100x development, and I'm doing that with confidence, not just like vibing through it, I think that will be a game changer moment for us, it means that we totally broke the way we work, we're going to get to that, so I think like, I differentiated between, I separated between noobs and enterprise, for example, I think for noobs, it means end-to-end means, for example, that you prompt and get a software, a fully managed software for you, right, we, I guess we're all playing with, even if you're a developer, you're doing some POVs, or some side, or being able to do some simple games, but then in the enterprise setup, in the heavy-duty setup, then basically, what means for us, like, AI cross-desk, you'll see as reliable, versatile workflows, and when I say workflows, we had a few sessions, amazing sessions, I'm actually, like, combining them, which is one of the points that were made there, like, I think a child made a point where an agent is creating workflows and calling them, and then there was a discussion about, like, workflows and, you know, using agent, et cetera, so I'm, when I say workflows, I already mean agentic workflow, let's, let's move on, like, workflow, agent, who cares, like, like, so it's going to be combined, and I think, like, in order to do that with trust, you need to be able, more importantly, is to have workflows or agents that are related to trust, for example, having really high-quality reviewing capabilities, really high-quality testing capabilities, being able to do a variety of jobs to be done that we do as developers in end-to-end-ish and high-quality, so that will be a game-changer.

Let's be a bit more specific. As developers, we do a lot of things, not just write code. I took, like, four parts, which is, like, I think less than a half of what we do, and I think, like, basically, vibe coding was very focused on, let's, like, do initial planning and write code fast, like, and, you know, vibe, whatever we get, like, that implementation or that, we feel it and move on, but actually, in heavy-duty software, we need to do many other things.

We need to do, for example, bug fixing, or we need to do refactoring and changing features, etc. We want to move fast without breaking things, so I think we need to have not the testing and the writing as a hand site, which you can do, like, testing and reviewing inside ID tools that are very mature, but not as a hand site, but rather, like, really being the red team, right, thinking as a tech lead, as reviewing code and having, like, fitting technology for that, so if we are able to put, like, AI in each one of those, like, parts, maybe then it will be a game-changer, but I think even then, like, we already have tools for code review, for testing, shameless plug, we're going to talk about later, that's what we do at Codo.

We focus being more on the red team, even though we have tools for the blue, we are focused more on the red team, more on the red team. Still, I don't think it's a game-changer, okay, and I think, like, a first, like, a feeling of a game-changer, as I mentioned, not only putting AI across the SDLC, but having workflows across parts of the SDLC, so one semi-game-changer could be, like, if we manage to shift left the review and the testing, as you write code, then you also get those suggestions and get high-quality code that is already according to your company best practices, and you don't need to wait for the AI review tool to tell you that only after you pushed it, right?

But I think, like, I think we have even a better way to do that, a bit more agentic, so let's reanalyze these four blocks, these four tasks, these four tasks, so I, instead of having one, like, the x-axis, now there's also y-axis, it's executability. If you know the v-shape in Wikipedia or whatever, where we're searching now, chat GPT, v-shape, software development, y-axis executability, x-axis time, so we start with planning, it's not executable, then we write code, in most cases, you're not doing admitted like a TDD, okay?

So you write, you write your code, and then you write your test, and in most cases, and that's executable, but it's not application code, so it's a bit higher, and then, and then you do the review. A game changer is, like, when we actually manage to take that v and squeeze it, and, and have, like, different workflows and agents talking to each other, and that's why, like, I had the word MCP and A2A, and I'm gonna get to that later, I, I, I guess, like, if you're in this conference, you know what I'm talking about, that where maybe there's different agents that are talking to each other, maybe it's one agent using tools that are able to connect all of these, like, all at once, that, that will be a game changer.

So to, and to some extent, what it means is that, that, like, the game changer moment is when we're looking at all of these steps, like, rigorously, and, and one holistic solution. If you like, cloud is, is mature for 10 years, and then Wiz came and said, we can't put in a security application for each part of the cloud, like, or each, uh, necessity of, of security.

Let's build one holistic, uh, solution for, for, for cloud security. The same thing will happen to us in the game changing moment for, for, uh, intelligent software development. Zooming out back again to vibe coding with confidence, uh, it, it didn't happen a long time ago when Karpathy wrote about vibe coding, like, you know, really, like, I thought it's like a decade already in AI, uh, years, uh, and, and the interesting thing, it didn't take too much time, like, a month or two, uh, where, where we started, where, where we started seeing pushbacks, right?

This is not me, by the way. Uh, and, and, uh, and, and actually just a few weeks ago, the same Karpathy came back and rethought vibe coding. Notice what he's saying, like, code I actually and professionally care about contrast vibe code. The same Karpathy, okay? And it's really interesting, I think, to analyze what he's saying.

The first thing he's saying, okay, coded I care about, probably have a duty software, or I like the quality, it's maintainable. We are gonna get to that. And then he tries to offer solution. I, uh, I love him. I, I, I mean, like, I never met him, but, but, uh, we all saw his videos, right?

Like, 15 years ago, 10 years ago. So, uh, he's giving solution, not only, like, uh, like, being annoying, like, okay, no, I said by coding, no by coding anymore. So, first he's saying, the first number one is, is context. He's saying, like, bring context. Here, he's saying doing it manual, but that's not game changer if now I need to start bringing it manually.

And then he's giving also two, three, four, five, six, seven. Like, this, this is a workflow. He's actually suggesting a workflow. That's one of the topics of the three talks that we had now, okay? So, now, uh, like, this is really important, I think. Like, how do you vibe code with confidence?

You need to, like, you said, it's like, just vibe coding, probably you don't care about quality. You do care. You need to bring the right context and have workflows. And these workflows can be connected to agents, could be part of the agent, could be part of your flows. And that's, like, in the nine minutes or so, some, some things I'm going to talk about.

Karpati and others are not the only one. This is just, like, from a week ago, uh, if you, you know, this, uh, a very, uh, great, great person, great material. He's saying, like, AI coding tools are so much less useful on existing large code bases, okay? He's saying that, and, and this is differently than compared green for greenfield work.

And it's not surprising because if you have a greenfield work or it's a simple software, code generation is all you need, okay? Like, it's magnificent, the tip of the iceberg, give a prompt, you get your software. But when you're talking about enterprise, heavy-duty software, then there's much more than that.

We all know it, right? There is a maintainability, testing, reviewing, et cetera. And shameless plug, well, I am the CEO of Koto, I have to say, we are looking at that as a holistic, like, how could we really bring a game-changer moment as if we look at all of this at a holistic approach?

That's why, uh, we have a multi-agent, uh, architecture. I am not going to go over it one by one because there's one specific thing that I wanted to show. So, we have, uh, a tool, we have an agent that is focused on, let's call it deep research, deep ask, in code.

And then, uh, we have a tool that is specific, that's our flagship for, for code review, called code merge. And then, finally, we have, uh, like, a tool that is inside IDEs that enable to shift left, like, bring that context, bring that, uh, bring that best, best practices that were collected over time.

The code merge tool, the code review tool collects best practices over time. And this, the context that is coming from the code aware and the best practices coming for code merge, helping you rather whatever code generation tool you're using or using directly code tool to make sure that that's in high quality.

But recently, like 20 minutes ago, uh, we released also a CLI tool. And that is meant to help you with the workflow parts. The workflow is already integrated into our solution, but this, like, really, uh, a workflow on steroids. There's another thing very important that is coming out in two months, and stay tuned, follow us, whatever, any platform that you want.

I kind of, like, uh, didn't want to show it right now. So, CLI tool, why? I hope, like, I gave enough, like, hints, um, and, uh, so far. And the main reason is putting AI to work across the CLC. And I remind you that starting five minutes ago, you can go download it and play with it.

It's a soft launch. Uh, we're going to announce about it more deeply, like, in, in, in a couple of weeks. So, interestingly, who, who, who saw Simon, like, uh, talk, like, he's magnificent? What? Some people didn't? Okay. So, please go, uh, go out of the room, I mean. And, uh, and then come back after and watch, uh, online later.

And it's amazing, like, when he wanted to demonstrate something, did he open, like, one of the IDEs? He did not. Like, everything he did was with, with a CLI tool. Like, that's, if you are a developer, like, that's the idea that you can run it in the background, you can run it in different workflows, you can make workflows.

That's part of what I'm, what I'm going to show you. And, and basically, with, with Codo, uh, the CLI tool, which is live, uh, live here, uh, then I can, for example, chat with it. And very interestingly, that is very, quite, like, there are a few tools that we talked about before that you can do that.

But interestingly, you can call agents. For example, I can call the, the explain agent. It's a dedicated agent. Uh, due to lack of time, I'm not going to show that. I'm going to show something else. So, how does, how did we made this explain agent? You can actually, with Codo the agent, you can make other agents.

And soon, I'm going to show you how you also create workflows and things like that. So, basically, I can go and ask it, if the internet works, uh, well enough, uh, you can go and ask it, please create an agent with this goal. And, and by the way, you can, uh, uh, uh, like, give much more information than what I'm doing right now.

I'm asking for it to create a review agent. Okay? If the internet works, it's, it's going to start working. It's initializing MCP. It's going to search what it can do, what it cannot do. And now it's, it's going to work for, for a few minutes. Okay? Like, if I would ask you to do it, you would do it, take it in two hours.

So, let's give it two minutes. Meanwhile, I'm going to, to progress and, and, uh, and tell you that the, I think one of the interesting things, uh, about it, it's, it's working. So, one, uh, most agents, like, I can't think of, of, of, of one that won't, would have tools.

Uh, for example, the Kodo cover agent, uh, by the way, we have an open source for Kodo cover. Uh, it's a tool that automatically, uh, generates coverage, coverage, uh, it increases the coverage, automatically increases coverage, but you don't need anymore to go download this open source, et cetera. You just get it as part of the, uh, uh, as one of the agent.

Okay? So, uh, but if you do want to play with the open source, et cetera, it's, it's fully open source. Okay. Back, back to my point, I can go and call Kodo cover, which uses different tools. For example, it uses tools related to coverage reports. Uh, and, and you, I can look what I can do in a command line.

I can tell it dump all the logs somewhere, right? How do you do that with an ID plugin? You can, but you need to hack it, hack it. But the interesting thing, what can we do with command lines? And many here are, are developers. You can pipe, you can pipe, uh, uh, tools together.

Agents, you can pipe, like, uh, uh, uh, workflows, et cetera. So I want to run Kodo code, generating code. And then I want to run the cover that, like, you know, increase the coverage testing. Maybe the code will generate a few tests, but I have an agent that is that is specializing in increasing coverage.

And then I want the review agent that is now cooking and finished. Uh, um, and I'll go back to it in a minute if I have time. Now, this is not no agent to engine communication. I did a pipeline that just passed the result, but what if we can, and, and, and, and before I exemplified with MCP, but what if we can invent a command line for A to A?

So those agents are, uh, uh, will run in parallel as you go and they will have some communication, uh, between them. So I actually have to say that yesterday we had like a 20, uh, 50, uh, like people like, like those that you had, uh, we, uh, uh, speakers and I asked who is using A to A?

No one is using it. So it's still a way to go, but this is an example. Oh, yes. Come to talk to me later. So I will know one person. Okay. Thank you. So, uh, uh, uh, and then I'll edit all the posts. Uh, I know one person and, and, uh, and, and, and, but, but we're getting there.

Okay. This is an example. You need multiple agents. You need A to A. You, you would want to use it if you want discover, discoverability between agents or handshakes. So this is a good example if you have many agents and you can run them in the background, et cetera.

So in this case, the agent finished and I have the definition. Uh, I have the definition of the, of the agent, uh, over, over here. Uh, it's called merge agent and you can see like, like all everything, the instruction that was made for it, uh, like what tools it is, it is using.

Uh, I'm going to add one that, uh, sometime I want to be very clear. It's there. And by the way, it also has like a, uh, output scheme. I'm going to remove that, move that for a second. Okay. And then I can go back, uh, and, and just call it, it was called merge review, uh, Kodo merge, uh, review agent.

It was just generated a minute ago and it will review. I didn't give it any, you see that there's a parameter automatically created like, uh, uh, that, that, uh, uh, I, I can tell it what to review, et cetera, give instruction. It's very flexible, but I didn't. Uh, okay.

So, uh, I'm almost lack of time, so I won't, uh, show you, but what I just will tell you that basically I can tell it to use best practices from the organization important for vibe coding with confidence. And I can also tell it to have some success or, or, or failure criteria to know that something happened well or not.

For example, if it managed to get some coverage, okay, that's how you work with, with confidence. Um, yeah. And you can connect things to pre-commit, post commit, okay. And everything you can do, you can do automatically. Very interestingly, and, and with that, uh, almost reached the end, very, very interestingly, you can say the CLI tool is very lame.

But notice, did you notice what, what Simon did? He used the CLI, but eventually checking content, he, he used different interfaces. So, you can, from the CLI itself, automatically generate interfaces that are fitting for your task. And one of those, for example, is an interface for, for code review.

So, who cares anymore about IDEs when you can have flexible IDs? And with that, I'll finish, and I'll just tell you that I think we're going into a future where we're going to have swarm of agents, each one specializing and having different credential, different, uh, best, best practices, etc.

And, uh, and, and, and it's coming really fast, 2025, 2026. Thank you.