Your agents lack context: Here's how to fix "You're absolutely right!" — Brandon Waselnuk, Unblocked

>> Good afternoon. >> Good afternoon. I hope you're all having a lovely day I hope you're all having a lovely day I hope you're all having a lovely day here at AIE. here at AIE. here at AIE. We've had great weather, though the UV We've had great weather, though the UV We've had great weather, though the UV has been like nine. So, hopefully you has been like nine. So, hopefully you has been like nine. So, hopefully you put your sunscreen on your being put your sunscreen on your being put your sunscreen on your being appropriate adults. appropriate adults. appropriate adults. I'm here to talk to you about context I'm here to talk to you about context I'm here to talk to you about context engineering, and I have the good fortune engineering, and I have the good fortune engineering, and I have the good fortune of following AJ from LinkedIn because he of following AJ from LinkedIn because he of following AJ from LinkedIn because he talked a lot about the system that we talked a lot about the system that we talked a lot about the system that we actually design and sell to other actually design and sell to other actually design and sell to other solutions. And I'm going to give you a solutions. And I'm going to give you a solutions. And I'm going to give you a bunch of open source tools. So, if you bunch of open source tools. So, if you bunch of open source tools. So, if you watch that last talk just before me, watch that last talk just before me, watch that last talk just before me, you're going to get a bunch of tool you're going to get a bunch of tool you're going to get a bunch of tool chance you can go mess around yourself, chance you can go mess around yourself, chance you can go mess around yourself, and I'll teach you a bunch of techniques and I'll teach you a bunch of techniques and I'll teach you a bunch of techniques today. The goal, of course, is to fix today. The goal, of course, is to fix today. The goal, of course, is to fix your absolutely right. your absolutely right. your absolutely right. I think they've taken that out of the I think they've taken that out of the I think they've taken that out of the prompts now, so it just says you're prompts now, so it just says you're prompts now, so it just says you're right or other things, but I'm sure right or other things, but I'm sure right or other things, but I'm sure you've all been there. you've all been there. you've all been there. So, I'm Brandon. So, I'm Brandon. So, I'm Brandon. I work at Unblocked. Uh yes, I have a I work at Unblocked. Uh yes, I have a I work at Unblocked. Uh yes, I have a coconut. We've been giving these away coconut. We've been giving these away coconut. We've been giving these away for fresh context, fresh fresh coconuts. for fresh context, fresh fresh coconuts. for fresh context, fresh fresh coconuts. But, the thing that I want to talk to But, the thing that I want to talk to But, the thing that I want to talk to you about is with these models, you about is with these models, you about is with these models, especially with Meth O'Clock models, I especially with Meth O'Clock models, I especially with Meth O'Clock models, I think Fable 5's coming back today, so think Fable 5's coming back today, so think Fable 5's coming back today, so they say. they say. they say. You can watch my Grain Call recording You can watch my Grain Call recording You can watch my Grain Call recording try to book this. try to book this. try to book this. We'll ignore it. We'll ignore it. We'll ignore it. But, what I want you to do is to think But, what I want you to do is to think But, what I want you to do is to think about the fact that with these tools, about the fact that with these tools, about the fact that with these tools, AI-generated code should feel like it AI-generated code should feel like it AI-generated code should feel like it was written by someone who's been on was written by someone who's been on was written by someone who's been on your team for years.

So, to get in the right headspace, for So, to get in the right headspace, for years you have to consider that you have years you have to consider that you have years you have to consider that you have been the context engine. been the context engine. been the context engine. How did you do that? How did you do that? How did you do that? You built context by going to work You built context by going to work You built context by going to work and asking questions, and asking questions, and asking questions, shipping PRs and getting them rejected, shipping PRs and getting them rejected, shipping PRs and getting them rejected, going to meetings, and all this slowly going to meetings, and all this slowly going to meetings, and all this slowly over time built up the engine that is over time built up the engine that is over time built up the engine that is your brain. your brain. your brain. You understand how it works here. You You understand how it works here. You You understand how it works here. You know how stuff gets shipped. You were on know how stuff gets shipped. You were on know how stuff gets shipped. You were on call that night when you took prod down call that night when you took prod down call that night when you took prod down and why that happened. and why that happened. and why that happened. The problem is The problem is The problem is that these agents have this exact same that these agents have this exact same that these agents have this exact same problem. Every time you create a new problem. Every time you create a new problem. Every time you create a new terminal session with an agent in it, terminal session with an agent in it, terminal session with an agent in it, it's very intelligent, but it doesn't it's very intelligent, but it doesn't it's very intelligent, but it doesn't have any context on how your company have any context on how your company have any context on how your company operates. So, it needs to get that operates. So, it needs to get that operates. So, it needs to get that somehow. somehow. somehow. The problem is as you move these agents The problem is as you move these agents The problem is as you move these agents up in scale, up in scale, up in scale, that cost compounds if you get it that cost compounds if you get it that cost compounds if you get it incorrect at the beginning. The leverage incorrect at the beginning. The leverage incorrect at the beginning. The leverage of context and content of context and content of context and content We're just going to fix this cuz I think We're just going to fix this cuz I think We're just going to fix this cuz I think people want to take some photos. Perfect. Perfect. That context issue will compound. So, at That context issue will compound. So, at That context issue will compound. So, at the far left, we all remember the the far left, we all remember the the far left, we all remember the age-old time of 2 years ago where we had age-old time of 2 years ago where we had age-old time of 2 years ago where we had tab complete models that were pretty tab complete models that were pretty tab complete models that were pretty cool. What happened is it popped up and cool. What happened is it popped up and cool. What happened is it popped up and said, "Hey, do you want to tab this?"

said, "Hey, do you want to tab this?" said, "Hey, do you want to tab this?" And quickly in your head with your And quickly in your head with your And quickly in your head with your context engine, you go, "No, that's context engine, you go, "No, that's context engine, you go, "No, that's bad." Or you went, "Oh, sweet." You hit bad." Or you went, "Oh, sweet." You hit bad." Or you went, "Oh, sweet." You hit tab. Nice. tab. Nice. tab. Nice. As we move along the agentic kind of As we move along the agentic kind of As we move along the agentic kind of adoption curve, what happens is you are adoption curve, what happens is you are adoption curve, what happens is you are moving into more situations in which you moving into more situations in which you moving into more situations in which you have agents running without a human in have agents running without a human in have agents running without a human in the loop, or at least you wish you the loop, or at least you wish you the loop, or at least you wish you didn't have to be in the loop. didn't have to be in the loop. didn't have to be in the loop. What they need is some way to be able to What they need is some way to be able to What they need is some way to be able to ask the questions they need when they ask the questions they need when they ask the questions they need when they hit walls in order to write code or hit walls in order to write code or hit walls in order to write code or solve or basically fix the issue and solve or basically fix the issue and solve or basically fix the issue and ultimately output code that's mergeable ultimately output code that's mergeable ultimately output code that's mergeable into your code base, especially with into your code base, especially with into your code base, especially with many people here who actually work in many people here who actually work in many people here who actually work in brownfield code bases that have been brownfield code bases that have been brownfield code bases that have been around for a long time that run real around for a long time that run real around for a long time that run real revenue across them, not just greenfield revenue across them, not just greenfield revenue across them, not just greenfield fun projects. fun projects. fun projects. So, that cost of bad context compounding So, that cost of bad context compounding So, that cost of bad context compounding at the beginning is cheap. If you think at the beginning is cheap. If you think at the beginning is cheap. If you think like shift left, finding a defect or a like shift left, finding a defect or a like shift left, finding a defect or a bug, you want to find it as early as bug, you want to find it as early as bug, you want to find it as early as possible. It's the same with context. possible. It's the same with context. possible. It's the same with context. Cuz as you move across, you get into Cuz as you move across, you get into Cuz as you move across, you get into doom loops. You usually ask your to do doom loops. You usually ask your to do doom loops. You usually ask your to do something. It's like, "Hey, I did it." something. It's like, "Hey, I did it." something. It's like, "Hey, I did it." And you're like, "No, man." And then you And you're like, "No, man." And then you And you're like, "No, man." And then you correct and correct and correct. That's correct and correct and correct. That's correct and correct and correct. That's wasted search tokens. It's also wasted wasted search tokens. It's also wasted wasted search tokens. It's also wasted rework time.

rework time. rework time. And that is not acceptable with the And that is not acceptable with the And that is not acceptable with the tokenomics we have coming. tokenomics we have coming. tokenomics we have coming. And then as you move into parallel And then as you move into parallel And then as you move into parallel agents, etc., you start hitting a review agents, etc., you start hitting a review agents, etc., you start hitting a review tax. So, these AI code reviewers we're tax. So, these AI code reviewers we're tax. So, these AI code reviewers we're trying to use, but again, key context is trying to use, but again, key context is trying to use, but again, key context is important there so that those code important there so that those code important there so that those code reviews are able to basically reviews are able to basically reviews are able to basically understand how the operations of the understand how the operations of the understand how the operations of the business are so it knows the business business are so it knows the business business are so it knows the business logic and more. logic and more. logic and more. And then finally, if your hope is to And then finally, if your hope is to And then finally, if your hope is to move all the way out of the loop, you're move all the way out of the loop, you're move all the way out of the loop, you're like background agents, get it done, like background agents, get it done, like background agents, get it done, make no mistakes, you really need to make no mistakes, you really need to make no mistakes, you really need to make sure that you have a context engine make sure that you have a context engine make sure that you have a context engine so those agents can query it and get all so those agents can query it and get all so those agents can query it and get all the answers they need so they can keep the answers they need so they can keep the answers they need so they can keep operating in an effective way. There are some common approaches that There are some common approaches that don't work. They're basically like a don't work. They're basically like a don't work. They're basically like a local maxima. local maxima. local maxima. Two of the ones we see the most with our Two of the ones we see the most with our Two of the ones we see the most with our hundreds of enterprise clients and hundreds of enterprise clients and hundreds of enterprise clients and mid-market size businesses mid-market size businesses mid-market size businesses is the curated context trap. If you've is the curated context trap. If you've is the curated context trap. If you've ever sat down and taken a virtual file ever sat down and taken a virtual file ever sat down and taken a virtual file system or maybe a local file system, you system or maybe a local file system, you system or maybe a local file system, you put some markdown files in it and you're put some markdown files in it and you're put some markdown files in it and you're like, here's all the context of this like, here's all the context of this like, here's all the context of this project, it's how it works. You then project, it's how it works. You then project, it's how it works. You then allow your agent to grep over that and allow your agent to grep over that and allow your agent to grep over that and it gets a bunch of good data and then it it gets a bunch of good data and then it it gets a bunch of good data and then it will perform better.

will perform better. will perform better. The issue is first, now you have to The issue is first, now you have to The issue is first, now you have to distribute that so maybe you throw it up distribute that so maybe you throw it up distribute that so maybe you throw it up in a GitHub and your team can grab it. in a GitHub and your team can grab it. in a GitHub and your team can grab it. But then the next is that repo is going But then the next is that repo is going But then the next is that repo is going to rot just like all the other docs you to rot just like all the other docs you to rot just like all the other docs you wrote down and then who at your org is wrote down and then who at your org is wrote down and then who at your org is the omnipotent one who has the taste to the omnipotent one who has the taste to the omnipotent one who has the taste to curate this file or repo for literally curate this file or repo for literally curate this file or repo for literally everyone in the org. So you start to hit everyone in the org. So you start to hit everyone in the org. So you start to hit these issues. these issues. these issues. The next is the MCP plateau. The next is the MCP plateau. The next is the MCP plateau. This one is pretty clear. We have MCPs, This one is pretty clear. We have MCPs, This one is pretty clear. We have MCPs, they're great. You can give it to your they're great. You can give it to your they're great. You can give it to your agent and now it can basically get agent and now it can basically get agent and now it can basically get information from another source system. information from another source system. information from another source system. The problem is, of course, based on how The problem is, of course, based on how The problem is, of course, based on how you write the server description, the you write the server description, the you write the server description, the tool descriptions, your agent may never tool descriptions, your agent may never tool descriptions, your agent may never call it even though it should have. Or call it even though it should have. Or call it even though it should have. Or if it does, there's a known bias called if it does, there's a known bias called if it does, there's a known bias called the satisfaction of search bias. What the satisfaction of search bias. What the satisfaction of search bias. What that means is the agent, when it finds that means is the agent, when it finds that means is the agent, when it finds the first piece of information that it the first piece of information that it the first piece of information that it thinks is correct, it goes, "Oh, I have thinks is correct, it goes, "Oh, I have thinks is correct, it goes, "Oh, I have what I need." and it proceeds. what I need." and it proceeds. what I need." and it proceeds. In most organizations, there's a Slack In most organizations, there's a Slack In most organizations, there's a Slack conversation from last night that says conversation from last night that says conversation from last night that says you should be doing A instead of doing B you should be doing A instead of doing B you should be doing A instead of doing B and the agent will never find it if it and the agent will never find it if it and the agent will never find it if it found some architecture record first.

found some architecture record first. found some architecture record first. So it doesn't actually consider all of So it doesn't actually consider all of So it doesn't actually consider all of the context. the context. the context. The problem here is access to The problem here is access to The problem here is access to information is not understanding. information is not understanding. information is not understanding. So to deliver understanding to a model, So to deliver understanding to a model, So to deliver understanding to a model, you have to do other techniques. you have to do other techniques. you have to do other techniques. What I'm basically trying to say is What I'm basically trying to say is What I'm basically trying to say is what your agent can't see is everything what your agent can't see is everything what your agent can't see is everything below the waterline. It can 100% get below the waterline. It can 100% get below the waterline. It can 100% get code that compiles, but that code that code that compiles, but that code that code that compiles, but that code that compiles is taking down prod and you compiles is taking down prod and you compiles is taking down prod and you have a P0 at 1:00 in the morning. have a P0 at 1:00 in the morning. have a P0 at 1:00 in the morning. Because it missed the fact that you have Because it missed the fact that you have Because it missed the fact that you have a certain rollout procedure, you're a certain rollout procedure, you're a certain rollout procedure, you're supposed to turn off a feature flag, supposed to turn off a feature flag, supposed to turn off a feature flag, whatever it might be. So, your team needs a context engine So, your team needs a context engine because what it should do is understand because what it should do is understand because what it should do is understand who you are and where you work in an who you are and where you work in an who you are and where you work in an organization. So, if I say to you, I organization. So, if I say to you, I organization. So, if I say to you, I want to get off stood up, it knows where want to get off stood up, it knows where want to get off stood up, it knows where I work, it knows where my get commits I work, it knows where my get commits I work, it knows where my get commits are, it knows who reviews those commits, are, it knows who reviews those commits, are, it knows who reviews those commits, and it understands that my context, it and it understands that my context, it and it understands that my context, it can focus me, and then use that as a can focus me, and then use that as a can focus me, and then use that as a trigger point to find the rest of the trigger point to find the rest of the trigger point to find the rest of the information. information. information. It resolves conflicts, as mentioned, an It resolves conflicts, as mentioned, an It resolves conflicts, as mentioned, an old architecture diagram and last old architecture diagram and last old architecture diagram and last night's Slack convo with the CTO, night's Slack convo with the CTO, night's Slack convo with the CTO, which one is right? You need to use a which one is right? You need to use a which one is right? You need to use a bunch of techniques to discern determine bunch of techniques to discern determine bunch of techniques to discern determine that.

that. that. Respects permissions and governance, of Respects permissions and governance, of Respects permissions and governance, of course. MCP allows us to use OAuth and course. MCP allows us to use OAuth and course. MCP allows us to use OAuth and other scopes and SSO, but if someone other scopes and SSO, but if someone other scopes and SSO, but if someone asks a question over here who's not asks a question over here who's not asks a question over here who's not supposed to know about secret project A, supposed to know about secret project A, supposed to know about secret project A, you need to make sure that doesn't leak you need to make sure that doesn't leak you need to make sure that doesn't leak into the response. into the response. into the response. And then finally, deliver the right And then finally, deliver the right And then finally, deliver the right context at the right time to the model context at the right time to the model context at the right time to the model in a token optimized way. in a token optimized way. in a token optimized way. We have multiple surface areas because We have multiple surface areas because We have multiple surface areas because human engineers still talk to Unblocked human engineers still talk to Unblocked human engineers still talk to Unblocked all the time to get information they all the time to get information they all the time to get information they need in Slack or otherwise, but then you need in Slack or otherwise, but then you need in Slack or otherwise, but then you want token optimized responses if you're want token optimized responses if you're want token optimized responses if you're just speaking machine to machine in just speaking machine to machine in just speaking machine to machine in order to not waste a bunch of bold order to not waste a bunch of bold order to not waste a bunch of bold classes on your token spend. classes on your token spend. classes on your token spend. This is how an engine works. I'm going This is how an engine works. I'm going This is how an engine works. I'm going to be brief on this, but basically on to be brief on this, but basically on to be brief on this, but basically on the left-hand side, the left-hand side, the left-hand side, you see all the data sources that are you see all the data sources that are you see all the data sources that are coming in. coming in. coming in. For us, we focus on engineering teams For us, we focus on engineering teams For us, we focus on engineering teams and that's who uses us, as well as the and that's who uses us, as well as the and that's who uses us, as well as the technically light teams around it, like technically light teams around it, like technically light teams around it, like support, sales, and otherwise. support, sales, and otherwise. support, sales, and otherwise. You ingest all that data, you get You ingest all that data, you get You ingest all that data, you get real-time data from tools like your real-time data from tools like your real-time data from tools like your instant management tool chain. instant management tool chain. instant management tool chain. It comes into the engine, where that It comes into the engine, where that It comes into the engine, where that engine is, it thinks at the bottom. I'll engine is, it thinks at the bottom. I'll engine is, it thinks at the bottom. I'll expand on that slide in a moment. But expand on that slide in a moment. But expand on that slide in a moment. But basically it uses these six key basically it uses these six key basically it uses these six key characteristics. And then on the right, characteristics. And then on the right, characteristics. And then on the right, you output the context to the exact you output the context to the exact you output the context to the exact workflow in the manner that it is workflow in the manner that it is workflow in the manner that it is needed.

needed. needed. Those six key points, as mentioned, Those six key points, as mentioned, Those six key points, as mentioned, unified system context, you have to go unified system context, you have to go unified system context, you have to go across the whole thing. At large orgs, across the whole thing. At large orgs, across the whole thing. At large orgs, companies like LinkedIn scale, Workday, companies like LinkedIn scale, Workday, companies like LinkedIn scale, Workday, General Motors, whatever, they need this General Motors, whatever, they need this General Motors, whatever, they need this type of data. They need to understand type of data. They need to understand type of data. They need to understand everything that's happening. And Threek everything that's happening. And Threek everything that's happening. And Threek this morning actually talking about this morning actually talking about this morning actually talking about Fable coming out potentially later Fable coming out potentially later Fable coming out potentially later today, today, today, he mentioned that you need to actually he mentioned that you need to actually he mentioned that you need to actually provide a map and then let Fable provide a map and then let Fable provide a map and then let Fable discover the territory. The way to help discover the territory. The way to help discover the territory. The way to help confine that is making sure that these confine that is making sure that these confine that is making sure that these models have access to all of the models have access to all of the models have access to all of the context, because they will find your context, because they will find your context, because they will find your unknown unknowns. unknown unknowns. unknown unknowns. There are definitely things going on in There are definitely things going on in There are definitely things going on in your company that you're just unaware your company that you're just unaware your company that you're just unaware of, but would be really helpful for the of, but would be really helpful for the of, but would be really helpful for the task you're trying to do. task you're trying to do. task you're trying to do. That will move faster, but the targeted That will move faster, but the targeted That will move faster, but the targeted retrieval, you should be able to if you retrieval, you should be able to if you retrieval, you should be able to if you provide a link quickly, unfurl it, get provide a link quickly, unfurl it, get provide a link quickly, unfurl it, get that document back and move along. So, that document back and move along. So, that document back and move along. So, two tasks, deep research, go long, two tasks, deep research, go long, two tasks, deep research, go long, that's fine, but you also need speed that's fine, but you also need speed that's fine, but you also need speed when speed is required. Conflict when speed is required. Conflict when speed is required. Conflict resolution, we already talked about resolution, we already talked about resolution, we already talked about that, but one thing says do A, one thing that, but one thing says do A, one thing that, but one thing says do A, one thing says do B, who is right? says do B, who is right? says do B, who is right? Personalized relevance, who am I, where Personalized relevance, who am I, where Personalized relevance, who am I, where do I work, what am I working on?

do I work, what am I working on? do I work, what am I working on? That token optimization, making sure the That token optimization, making sure the That token optimization, making sure the response is good and effective and response is good and effective and response is good and effective and doesn't bloat the window. doesn't bloat the window. doesn't bloat the window. And then permission enforcement, of And then permission enforcement, of And then permission enforcement, of course, OAuth, you shouldn't see it, you course, OAuth, you shouldn't see it, you course, OAuth, you shouldn't see it, you shouldn't see it. shouldn't see it. shouldn't see it. What we did with some tests is we What we did with some tests is we What we did with some tests is we actually ran the exact same prompt to actually ran the exact same prompt to actually ran the exact same prompt to the same model and one with context and the same model and one with context and the same model and one with context and one without. This is the wall clock time one without. This is the wall clock time one without. This is the wall clock time savings. savings. savings. And then 2 hours, which is great. And And then 2 hours, which is great. And And then 2 hours, which is great. And then the tokens savings. So, it was a then the tokens savings. So, it was a then the tokens savings. So, it was a sizable task, it took about 21 million sizable task, it took about 21 million sizable task, it took about 21 million tokens without and then 18, or sorry, tokens without and then 18, or sorry, tokens without and then 18, or sorry, 10.8 million tokens with it. 10.8 million tokens with it. 10.8 million tokens with it. This is the type of experience that you This is the type of experience that you This is the type of experience that you typically see when you're using a typically see when you're using a typically see when you're using a context engine, cuz the majority of context engine, cuz the majority of context engine, cuz the majority of those wasted search tokens where it has those wasted search tokens where it has those wasted search tokens where it has to grab at the beginning of every to grab at the beginning of every to grab at the beginning of every session to understand and discover session to understand and discover session to understand and discover things are no longer there when it's things are no longer there when it's things are no longer there when it's hydrated with context. Hydrated. hydrated with context. Hydrated. hydrated with context. Hydrated. And then, as you move forward, you get And then, as you move forward, you get And then, as you move forward, you get these types of outcomes. these types of outcomes. these types of outcomes. 50% fewer tokens, faster triage, and the 50% fewer tokens, faster triage, and the 50% fewer tokens, faster triage, and the answer quality is actually better answer quality is actually better answer quality is actually better because it knew what was going on inside because it knew what was going on inside because it knew what was going on inside of the business. of the business. of the business. Now, this next part, Now, this next part, Now, this next part, you'll probably want to photo. If you you'll probably want to photo. If you you'll probably want to photo. If you don't know, you can actually take a don't know, you can actually take a don't know, you can actually take a picture of a QR code and then later in picture of a QR code and then later in picture of a QR code and then later in photos tap on it and then load the link photos tap on it and then load the link photos tap on it and then load the link so you don't need to float here cuz I'm so you don't need to float here cuz I'm so you don't need to float here cuz I'm going to give you three QR codes.

going to give you three QR codes. going to give you three QR codes. This first one is for the social comment This first one is for the social comment This first one is for the social comment network. I'll pop that up so you can network. I'll pop that up so you can network. I'll pop that up so you can take a photo. take a photo. take a photo. But, this is an open source tool that But, this is an open source tool that But, this is an open source tool that we've got that actually, using all we've got that actually, using all we've got that actually, using all deterministic programming, goes over deterministic programming, goes over deterministic programming, goes over your GitHub and understands who works on your GitHub and understands who works on your GitHub and understands who works on your team. This is my real team. We your team. This is my real team. We your team. This is my real team. We called Rasheem the machine cuz he ships called Rasheem the machine cuz he ships called Rasheem the machine cuz he ships like crazy. But, on the right, you can like crazy. But, on the right, you can like crazy. But, on the right, you can see who he commits, where he commits, see who he commits, where he commits, see who he commits, where he commits, who's reviewing his work. And then in who's reviewing his work. And then in who's reviewing his work. And then in those tabs, you can find a distilled those tabs, you can find a distilled those tabs, you can find a distilled experts graph. You get full coverage of experts graph. You get full coverage of experts graph. You get full coverage of what's going on in your business. And if what's going on in your business. And if what's going on in your business. And if you optionally add one of the API keys you optionally add one of the API keys you optionally add one of the API keys for either OpenAI or um Anthropic, it'll for either OpenAI or um Anthropic, it'll for either OpenAI or um Anthropic, it'll um determine what your teams are by um determine what your teams are by um determine what your teams are by doing some labeling for you. doing some labeling for you. doing some labeling for you. It's a really cool tool to understand It's a really cool tool to understand It's a really cool tool to understand where your team works and get that where your team works and get that where your team works and get that social network in there in order to social network in there in order to social network in there in order to focus the context engine if you're going focus the context engine if you're going focus the context engine if you're going to be building these tools yourself. to be building these tools yourself. to be building these tools yourself. The next is called the repo rules agent. The next is called the repo rules agent. The next is called the repo rules agent. This is a sample from our real code This is a sample from our real code This is a sample from our real code base. I'm going to pop that up anyway so base. I'm going to pop that up anyway so base. I'm going to pop that up anyway so you don't need to talk to the thing, but you don't need to talk to the thing, but you don't need to talk to the thing, but in short, what it does is discover all in short, what it does is discover all in short, what it does is discover all the places your team has written rules the places your team has written rules the places your team has written rules files, checks them all, and then tells files, checks them all, and then tells files, checks them all, and then tells you what severities you've given, you what severities you've given, you what severities you've given, what other things you've given. Should I what other things you've given. Should I what other things you've given. Should I just switch to this?

just switch to this? just switch to this? It tells you what it Whoa, hey. It tells you what it Whoa, hey. It tells you what it Whoa, hey. It's good to meet you all. It's good to meet you all. It's good to meet you all. Basically, it will find all the rules Basically, it will find all the rules Basically, it will find all the rules that are inside of your repo and then that are inside of your repo and then that are inside of your repo and then tell you if you have duplicate issues or tell you if you have duplicate issues or tell you if you have duplicate issues or others problems and then you can grab others problems and then you can grab others problems and then you can grab over it as an index. So, that index can over it as an index. So, that index can over it as an index. So, that index can be called and you can dedupe and it'll be called and you can dedupe and it'll be called and you can dedupe and it'll help improve um your retrieval of help improve um your retrieval of help improve um your retrieval of context. context. context. And then finally, on Monday we delivered And then finally, on Monday we delivered And then finally, on Monday we delivered this workshop, which was going beyond this workshop, which was going beyond this workshop, which was going beyond rag and taught how to build a relational rag and taught how to build a relational rag and taught how to build a relational context engine from scratch. context engine from scratch. context engine from scratch. So, if you scan that, you'll get the So, if you scan that, you'll get the So, if you scan that, you'll get the full workbook. It has six PRs stacked full workbook. It has six PRs stacked full workbook. It has six PRs stacked that teach you how to walk through doing that teach you how to walk through doing that teach you how to walk through doing this. But in short, rag is an incredible this. But in short, rag is an incredible this. But in short, rag is an incredible technique and you want that. But the technique and you want that. But the technique and you want that. But the other half of the problem is what people other half of the problem is what people other half of the problem is what people actually ask is, "What are the open PRs actually ask is, "What are the open PRs actually ask is, "What are the open PRs that I worked on in the last week with that I worked on in the last week with that I worked on in the last week with authentication?" authentication?" authentication?" Rag cannot answer that question alone. Rag cannot answer that question alone. Rag cannot answer that question alone. You need queries. So, this shows you how You need queries. So, this shows you how You need queries. So, this shows you how to do a to do a to do a schema-less basically look up that schema-less basically look up that schema-less basically look up that allows the agent to discover a schema allows the agent to discover a schema allows the agent to discover a schema and then write queries against it and then write queries against it and then write queries against it deterministically in order to get that deterministically in order to get that deterministically in order to get that type of relational data out.

type of relational data out. type of relational data out. Very useful technique. Use cases of a context engine, of Use cases of a context engine, of course, do go beyond code generation. course, do go beyond code generation. course, do go beyond code generation. This is, you know, where we live a lot, This is, you know, where we live a lot, This is, you know, where we live a lot, a lot of our customers spend their time. a lot of our customers spend their time. a lot of our customers spend their time. But it's amazing to see what happens But it's amazing to see what happens But it's amazing to see what happens when a bunch of other people around the when a bunch of other people around the when a bunch of other people around the business start picking up these tools, business start picking up these tools, business start picking up these tools, customer success people solving tickets customer success people solving tickets customer success people solving tickets right at the time that it comes in from right at the time that it comes in from right at the time that it comes in from a customer. a customer. a customer. We've got sales people closing deals We've got sales people closing deals We've got sales people closing deals earlier in their quarter because they're earlier in their quarter because they're earlier in their quarter because they're able to just query the Unblocked context able to just query the Unblocked context able to just query the Unblocked context engine on the fly while in the field. engine on the fly while in the field. engine on the fly while in the field. And so many more. What you can also do is if you saw that What you can also do is if you saw that curve chart earlier where I talked about curve chart earlier where I talked about curve chart earlier where I talked about the levels, we've built a fun little the levels, we've built a fun little the levels, we've built a fun little tool where basically an LLM will quiz tool where basically an LLM will quiz tool where basically an LLM will quiz you and ask you about what's going on you and ask you about what's going on you and ask you about what's going on and then it will map you to exactly and then it will map you to exactly and then it will map you to exactly where you are and then tell you some where you are and then tell you some where you are and then tell you some techniques about how to level up through techniques about how to level up through techniques about how to level up through that if you are looking to basically that if you are looking to basically that if you are looking to basically compound your capabilities and ship with compound your capabilities and ship with compound your capabilities and ship with AI tools at scale. It's AI tools at scale. It's AI tools at scale. It's readiness.unblocked.com. The gap is not intelligence any longer. The gap is not intelligence any longer. It's context. We will continue to get It's context. We will continue to get It's context. We will continue to get incredible models like Mythos as it's incredible models like Mythos as it's incredible models like Mythos as it's been grown by Anthropic and I'm sure been grown by Anthropic and I'm sure been grown by Anthropic and I'm sure Soul once I'm allowed to see it. I will Soul once I'm allowed to see it. I will Soul once I'm allowed to see it. I will get it. Happy Canada Day, by the way.

get it. Happy Canada Day, by the way. get it. Happy Canada Day, by the way. But what's happening is it's about the But what's happening is it's about the But what's happening is it's about the context you surround these models with context you surround these models with context you surround these models with in order for them to be effective and in order for them to be effective and in order for them to be effective and token efficient inside of your token efficient inside of your token efficient inside of your organization. So, I have a question slide, but I'm not So, I have a question slide, but I'm not sure I'm allowed. sure I'm allowed. sure I'm allowed. Nope. So, what you'll do is come meet me Nope. So, what you'll do is come meet me Nope. So, what you'll do is come meet me at booth P16. You can look for the at booth P16. You can look for the at booth P16. You can look for the coconut. coconut. coconut. It'll be great to hang out with all of It'll be great to hang out with all of It'll be great to hang out with all of you and get into details here if you you and get into details here if you you and get into details here if you need it. Thank you for your time. need it. Thank you for your time. need it. Thank you for your time. >> [applause]