GNG Analyst Walkthrough: AI Stock Research, Workflows and the War Room

Transcript

Introducing the GNG AI Console

One. All right, today we're just gonna show you briefly the GNG AI console, also known as the GNG Analyst. um, I'm also probably gonna just be walking through some new features with the team, uh, capabilities, why it's different, how it's different, and what you're able to do with this. So we're each just gonna go through and show you a couple of our favorite things. Um, first I'll show you one of my favorite things is the ability for our database layer to live on top of the AI system itself and fully integrate with your data.

So you'll have your portfolios that are connected, but also other external databases like all fundamentals, all price data, option data, the list goes on. Hundreds of tools, and it's, uh, the toolkit that AI is able to do. And also you can make your own skills and agents to help complete tasks.

SEC filings deep dive

Um, so first I'll show an SEC deep dive while I talk about it. So excuse my adolescent, uh, vocabulary here. So find business for added feed. Let's just go there and see what, uh, the output is. But essentially how it works is all documents that are uploaded in the SEC filings are immediately vectorized. And for the non-technical people out there, what that means is the AI just has the instant ability to search for values based on the meaning behind the actual values using some cool math and other features and functionality.

So right now, the AI is going through all this ticker's, Nvidia's, specific filings and trying to find anything that might have a meaning similar to its business strategy. And it's like this for every ticker within our database. Um, as you can see here, it even cites the sources, so it'll tell you, you know, what's the platform strategy or anything specific to that stock, obviously filings information. So that is the first thing I was

Workflows: scheduled and event-triggered research

going to show. But another example real quick, I wanna just introduce our workflows. I could drag and drop kind of a, a node style design like a lot of people are used to nowadays, or I could go directly into AI and say, "Create a workflow that emails me an update if anything changes in this system." And for our audience, Connor, what is a workflow? Good point. Glen, uh, is always the one to correct me when I'm a little bit too technical.

Someone's gotta keep you straight, Connor. So a workflow is essentially a triggered event or scheduled event. You can do time-based, like eight AM every morning, or a market triggered event. Hey, did the price drop by X amount? Did the dividend yield increase? Is there breaking news? And all this does is it goes through, and it triggers the beginning of a system, so the AI or whatever data flow that you integrate into our system begins. It'll prompt the GNG analyst and structure an output in a systematic way that you configure in the workflow settings or work with your agent live on the screen to configure, and then produce essentially your desired result.

So, you know, one example would be monitor my portfolio, since we do have brokerage integrated portfolios, monitor my portfolios, if any position drops five percent within a day. Do research into the price, do research into the news, um, research the fundamentals, thesis, earnings, earnings estimates or the revisions, find out what happened, and send me an email immediately. And so that way it just allows retail investors to go through and have curated personalized results that they build, right?

It's not recommendations or anything. It's, it's their research, but in an automated fashion. So you'll be able to monitor and find things that typically, you know, you may be at work or you may be busy, but your agent will always be in the background, whatever you configured to do to notify you whenever events occur, anything you feel like. So it's like having your own muse, but it's limited to just within this very private space, right? That's what I'm hearing.

And in finance, of course. Everything's related to what you're interested in, related to your stock investing or ETF investing. Exactly. And as you can see here, The AI agent went through and, and built out a designed workflow, so you didn't have to do any of the configuration, anything like that. It decided to use a trigger. It sends it to the AI analyst. It configures a model, settings, the prompt, what it's doing, uh, configured a saved report right here in your dashboard.

also configured a structured event, which can go through and see if, hey, did filings change, right? and then if it did, in this case, it'll go through, use the email writer and again, your structured format. It decided that SONNET was four point six was achieved. Good option. Uh, and obviously all of these are configurable with all the models we have in our system, and it'll email you. you can save these. You can activate and deactivate them.

You can also refine it, meaning the AI has memory of what the workflow is in the system and is able to go through and edit specific components, not just build new ones. But you can also test it live in the dashboard and go through, like test. It'll say running. It'll show you green lights as you go through. let me try to zoom in here a little bit. Green dots where things worked, and as the AI loads, it'll send you an automated report of the exact occurrence.

That is the first thing we wanna do. Also, you can optimize your credits, so on and so forth. But we won't wait around for the AI agent to finish the test itself. Glenn, I'll pass it off to you. Yeah, yeah. And what's really cool about what you just did is I use that actually almost every day, for finding when a stock ticker, um, has its momentum change. So you can do something as simple as a MACD, which is basically a, a momentum technical indicator, and you can say, like, if there's a MACD crossover, for anybody that's technical, you would know that, uh, go ahead and fire off an event that n-notifies me, uh, when that happens.

You can do this from fundamentals to technicals to news. it's

Skills, agents, memory, and the Rule of 40

really cool. So moving on, um, within our AI space, we have the ability to write our own sub-agents that do things or what we call a skill, and the difference is kind of subtle. Uh, a skill can be something that is a little bit more, uh, specific, so you think about specificity. And then the agent can be where it's a little bit heavier or beefier, and so it's broad and deep on what it's trying to accomplish.

And kind of we have agents that can call agents now too, right? We can spawn tasks that run 10 agents in parallel and things like that. Uh, so it's really cool. So I thought a good one to show because a lot of people wanna try and beat the S&P and are trying to find stocks with good growth opportunities, uh, is something called the Rule of 40. And it was really originally developed as a benchmark for the SAS financial space, and it's basically your growth rate plus your profit margin has to be at least 40%.

So Connor, if you could illustrate a growth, uh, Rule of 40, uh, example in Canon. And what he did is he actually typed slash, and then the skills, uh, will show which ones you can use. But the AI is smart enough that once you have this skill or an agent installed, all you have to do is type Rule of 40 and, and it knows, right? it knows to go use that skill. you know, you can always say, "Don't use this skill."

Uh, but, you know, that's the really cool thing is it has context. There's an entire context memory spot for what your account is, and it's private to you, that is used to kind of see how you like to do things and how you like to interact and of course everything that you install for agents and skills and y- your portfolio or portfolios. And what's really neat is, you know, every free user gets 100 credits. Like that SEC, uh, report took two credits, right?

So 100 credits, you could run 50 deep dive SEC, reports, which is pretty cool. While this loads, I wanna touch on memory for a second also. You can upload files directly. You can save, uh, very large documents. and we save that to your specific user profile and your context memory. Obviously, there's privacy settings. You can enable and disable what you would like. you're able to onboard yourself if you have a statement you wanna upload. if you have, you know, just your investing scope in general, you can take a PDF, a Word doc, upload it, and the AI is also able to invoke those memories as the context that you're talking with it is relatively similar.

it's similar to the skill set. it sees that you said Rule of 40, and it automatically invokes that skill, loading the context into the window dynamically instead of you having to do it every time. Yeah. In fact, I wrote an entire analysis engine, and all I did was upload MD files, Markdown files, and it just knows to go run it. So you can search for Glenn and Volkin or go to my, you know, some bio pages about me.

But, it fully works in here too, much less, uh, you know, within my own computer. Uh, so it's so powerful with the memory, so I'm glad you brought up memory,

Screen results and the research terminal

Connor. So What are we running here, so people understand, and what is it showing us? So, uh, he used a very specific example. So anytime you write an agent or a skill, the author, in this case it was me, uh, gives some examples, and we have an AI system that'll actually give you examples based upon how you wrote the skill, and this is actually one that the AI system offered that's within the documentation page in the marketplace.

And so we're basically looking in the USA with the Rule of 50, uh, and a cutoff of a $2 billion market cap. Awesome. And as you can see- You notice right at the top are some big boys, right? Micron and AppLovin, and I think AppLovin was-- I thought I saw them at the top. Maybe I missed it. Um, yeah, they're, no, number 12, and Nvidia is number 11. Awesome. And yeah, one of the most unique features is within the table, the feed itself, the results are shown, in a table format.

You can also click the See All button. and it'll take you directly to our research terminal where you'll see the exact screen that you-- was just on set up with all of our filters and, uh, the columns enabled exactly like the AI- You know what's really cool? Yeah, you could have asked right there on that research screen You can analyze AI from there, or you can just type in what you want and go through the filters and what columns, 'cause we have what?

Three hundred and ninety-one metrics that you could look at I think we're at five fifty-three hundred, uh, stock tickers now. And that includes ETFs, in our listings as well. So it's pretty awesome, right? that's the great thing about the AI system, but you can do this manually too, uh, kinda on your own and with your own tools if you'd like, 'cause you can export all this. But, you know, Connor's looking here at the top five you know, try and find the best 10 based upon, uh, backtesting it, right?

So that's an interesting test that, you decided to come up with. What is a backtest,

Backtesting and Monte Carlo forecasting

Connor? Yeah, I was about to explain this real quick because this is very unique to us. So our AI is able to actually invoke pretty much every tool we have in the system, and so it'll use our compute and our engines. Um, it is not an LLM sitting there trying to decide, where do I get the data? How do I run a backtest? We have an institutional level backtesting system or in forecasting suite that it'll go through and automatically send, what you asked it to that screen, get the result, and then live in the feed here, you'll be able to see the widgets pretty fast.

You'll see the, the widgets appear. You'll see the results, what you wanted to look at, what the test was. You can even click similarly to the last table I showed, Open in Backtester, right? And you can see the full, uh, run that was done and what the results were and, a ton of other data. And you get a history of it up in the top right. You can compare this particular one ticker to another one directly on the screen, and it'll show you side by side how they perform through all the regimes.

I mean, I'm sorry to interrupt you, Connor. I'm just so excited and the amount of, data that we give people is incredible compared to what I'm used to. And I think what's unique is that even the system itself will tell you, "Hey, if, if data was missing or something was weird, this is what happened, and this is what, how it was handled," right? So there's full transparency there, in terms of the system itself, and most of the handlings and stuff would be like, "Hey, this company IPO'd on such date, so the backtesting system handled it this way," right?

It wasn't out at the beginning of this period, um, and so on. And also similarly, our Monte Carlo, forecasting system. Monte Carlo's, on a financial level, is really good. They're really good at portfolio risk assessment. Um, for those of you who don't know what it is, it's just fancy math that does probability distributions, of a result. And our documentation includes how our exact tool works, our configurations

How Leo researches a stock: EPS outlook and estimates

are set up. And Leo, I wanted to ask you, from your use so far, what do you think is the most beneficial, that people would wanna know about? I think the most beneficial thing is that it's basically, a personal advisor. For example, we have so many tools on the website, um, and for my own convenience, I basically only use the AI tool and just tell it, um, just what you just did. Um, backtest these stocks for me, screen for this.

Um, and what I do a lot is, um, when I research companies, I wanna know specifically what's the outlook, for example. Then I, I ask, "Can you give me, the EPS outlook for the next few years for, for example," to name a stock, uh, GE Aerospace, ticker GE. Um, and I also ask it to provide data on how many estimates it is based on, like how many analysts. Um, that, that's what I use a lot, and then I tend to combine it with our advanced charting tools if I post something in an article or just, uh, you know, really quickly as- assess, um, the valuation.

And that's basically my starting point. Once I'm there and I have developed a thesis that requires a certain company, I basically scan the entire company, ask it about balance sheet health, um- Hey Leo, give me a ticker so I can run that in the background. Yeah, GE. And do you want current, uh, EPS revisions and how many analysts? Yeah, I would like to, like to know, you know, E- the EPS outlook for the next few years, um, the annual EPS outlook.

And, um, give me also how many analyst estimates this is based on. But sometimes you have a company where it's just based on one estimate. Yep. Awesome. And so, uh, Leo, one thing I'm really glad you brought up was our new advanced charting tool. I'll let you kind of dive deeper into that and discuss. Uh, I guess we'll talk about what it is at first, and then hopefully once this finishes, we'll, we'll link on over and we'll look at the chart so we, we can explain a little bit better.

Yeah, because then that, that's what I do as well. I just, in this AI chat, I just tell it, you know, "Give me the advanced charting tool," and it gives me a tool, and I can just click on Go to the Charting Tool. So we can do that in a second, and that's something I use almost, almost daily, I would say. Yeah. I, I love the, the advanced tool. I use it daily too. it's probably one of my top two tools.

Uh, obviously the research terminal and AI and the charting. Yeah, and it's so great to see that it was already included in the answer to the prompt. You just ask the AI. I saw when, when Connor scrolled through the answer that the advanced charting was included already. Um, so that is ve- very neat that you just ask it one thing and you just get the foundation. You can just pick whatever you need from a in-depth, analysis and just...

a few years ago you just had to read like 10 articles to get, um, to get the research you now get within seconds. even the system itself was able to determine GE today is GE Aerospace, and it described in, within the system that there's an old ticker that is General Electric. So it was able to see that-

Stock widgets and the advanced charting tool

we'll go through here. So you can see a ton of different widgets that we have whenever you just kinda ask for an overview or general things on a stock. So you can h- you have a price chart, right? You have your, uh, just a general valuation card or, Typical statistics, not really anything specific. If you want something super specific, it would then have to use our screening tool and go through and it'll show you a, a basically a table with all the metrics.

Um, our current internal scoring model and system, along with Wall Street dates and our quant system. Um, and of course, our new advanced charting tool and features. So then this is rendered directly here within the, chat, um, and includes a windowed summary based on the valuation and the metric multiple that

Crowdsourced estimates and valuation checks

is defined. And- Let's talk about that for a, for a second, though, since we're there. Because why is this so interesting to me is because unlike sell-side, uh, analysts, which would be, you know, your JPMorgan's of the world and Goldman Sachs, this is using crowdsourced information from one of the sources we use, which is FinHub. You know, it's, again, we're very transparent on everything, and they have something called Open Estimate. Uh, there was an open source group that put out something called Estimise that made this very popular, and FinHub has adopted that, and it's been proven by independent studies to be 60% more accurate on its values w- if you're within 30 days of a ticker's next earnings, report.

Uh, so to me, those are really valuable pieces of data. and it's still really good outside of that window, but the closer you get to the, the earnings window, the, the better probability that this, way of crowdsourcing your metrics for a company's growth and valuations, et cetera, is really powerful stuff. Because sell side is a, usually a lagger. You know, they're usually a couple of months behind, where the crowdsource has a lot more input fast. that's how I use it as well.

It's just, I mean, obviously, when you research a stock, that goes many, many hours into it, but this is an incredible foundation. You just look up the stock, you'll see where it trades now, and you see the earnings estimates. And of, of course, when you build the fees, you can say I'm more bearish or, or more bullish than the analyst. But this is exactly the quick and dirty, approach that, that gives you the foundation for further research.

this is something that I, I use probably on, on 10 to 15 stocks a day. I quickly scan them. I, I usually go out like four to five years, depending on, on the data. And immediately you see what's the valuation. is there a premium? Is there a discount? And, and usually when there is a big, big premium, we need to find out, you know, why is that? I mean, that's the ne- next big task. Why is there a premium, and is it worth it?

Or even if, if there is a discount, is it justified? Is, are there hidden risk? And that's when I go back to our AI again and then try to figure out why is it trading the way it does and just that's when you start to go into more debt. But this is, I think one of our best additions. I completely agree with Glenn here, I use it just as much as you. And if we can close the projection window, quite often what I find is I may not agree as an analyst because I, I can do a little bit deeper dive.

Uh, I might not agree with the formula's fair value. So we'll take this, an easy example that we have something called a P/E blended that's already been calculated. And just for, illustration purposes, let's change that 30 to 40, which is what the P/E blended is on the screen and see what happens, right? So now our fair value, values, which are down at the bottom where it says implied returns, it shows the custom fair value, and this is how much you can expect as annualized returns.

Right, guys? So that's really powerful information. I can kind of in a quick few seconds give you a lot at your fingertips to then do, do further research or make decisions on based upon this. Yep, and, and know whenever you go through this, I just always have the, gut feeling to say this, none of this is investment advice. we're providing you the tools to do whatever type of research, fits your exact niche, right? So, you know, if you're a fundamental investor, a long-term investor, something like this where you already understand components of that niche and, and charting like this would be super beneficial and save you some time, like Leo was saying, an at a glance view and, and really saving you even more time by interfacing it, with it through the AI, which can do the projection feature or, you know, the, the multiple inputs or change whatever metric you want to use for

the, in terms of valuation to fit that company. for example, the majority would be P/E ratio or REITs would be FFO, um, and these will appear dynamically. FFO does appear for REITs, if you were to actually look up a REIT, and that would be on the page. So that, that's just another quick example of, uh, the AI chat. Thank

Sentiment, insider trades, and ETF look-through

you, Leo. There, there's so many different features within this set and tools. I don't think we'll be able to go over every single one of them. Different charts, different, data points. For example, our market sentiment, gauge goes through and it'll, it'll tell you how many articles were analyzed over what period of time through our, our system, to say w- how the market feels about something. Uh, you'll be able to see different days or different scores. So you can also visualize the momentum that the market's going or even insider trading.

if it's a company that has government contracts, what is the trend of that? What's the history? That's included in here. if it's an ETF, what are all the holdings? Uh, you could ask the AI, "Look at my portfolio and all my ETF and individual holdings and tell me my true weight, of every single ticker that I own, even through ETFs." on the portfolio tracking page, though I don't want to go through any of my portfolios at the moment, um, there's a whole segment on that, right?

Our, our looking glass. So a very rich- System that everything under the hood really wires into the AI itself. So if you don't know if something's possible, ask the AI. It'll go through our documentation. if you wanna know how you can do something, ask AI. It'll go through the documentation and figure it out and perform it if it can for you, right? So that's just

The War Room: AI agents debate a question

a few things. Uh, final thing I'm gonna show here today would just be our war room. I'm gonna give a, a very brief run of how this works. it is one of one. we were one of the first to do it. Don't really know, um, of anyone else that was doing it at the time we were doing it. Still haven't seen a lot. You can go through and have the AI design a panel for you, and essentially your panels, all this is going to do is there's going to be a main agent that's going to allow other agents to argue whatever facts.

Uh, it has different settings, like if you want it to be hostile, if you wanna create a research team. They can use the sub-agents that you configure under agents, or it'll create its own agents to have it, discussed within the panel. So for time's sake here, I'm gonna go through and make it only two agents. We'll do hostile mode. these are all the different modes. You can also do sequential, which would just mean they talk to each other in terms of rounds or one after the other, or they can go all at the same time.

You can choose how many rounds there are, and you can also add a budget limiter, so it will not be able to surpass or spend a certain amount of tokens. everything else will just kinda let the agent run it for me. The orchestrator, uh, I clicked one of our preset options and had it design agents. And so while that runs, Glenn, any comments or, Leo, any comments on the war room in general, or any use case- Oh, yeah, I have one.

There's actually a skill. We haven't shown marketplace yet. Someone wrote a really cool skill called War Room CIO. You don't even have to come to this tool now, this particular page. Although this page is really nice, and so I do recommend using it. But you can run the War Room CIO skill w- from within the AI console, and it'll go off and do it. You just give it the prompt that you want, and it goes off and runs the war for you.

But I, I really find this useful. I'm a little indecisive between, let's say, two stocks. A classic example might be between Philip Morris and Altria if we're on the dividend side. If it's on the growth side, maybe we go with AMD versus Nvidia today. Like, if I was gonna put money into one of those two stocks today, which would I do? And they can either work as a group together, or they can try and challenge the other one to be, "I don't believe you're right, and here's why," with all those different settings.

So for me, it's kinda helps me maybe dig into things I didn't think about, I think is the best way to put it. 'Cause it'll go off and kind of do all this back and forth with ideas that I just d-- couldn't come up with, unless I'm gonna spend weeks researching it. But here you can do it. I mean, some of the war rooms, if you do multiple rounds, 'cause what's cool about the rounds, we only have it for one to make it quick, is it'll iterate through.

And so it'll be round one. It'll say, "Okay." And then the war room CIO, whoever your, your commander is that's running, at the main top of the war room, says, "Okay, we're gonna go to round two. Here's everything we've learned from everybody to take as input." You know, what do we still think is true within each of your spaces? Because you might, like, you hear-- see here you have a fundamental valuation analyst versus a quant and technical strategist.

These are completely different pieces, but they can work in tandem. Connor decided to have the fundamental team work against the, the quant and technical team, which by the way, happens all the time if you're on X or Reddit or anywhere on social media is, uh, you know, the technical people are very much so in their camp and the fundamentals in theirs, and sometimes there's conflict as to how you should be investing/trading. but I, I use both together because it's so powerful together.

Awesome. Thank you. And while this finishes up, um, you can see that these agents were actually determined to be different versions of models. So the DeepSeek is an option within our system, a very cheap model. it went through and used the pro version on one of them, and it used the flash version on the other. The flash is fast, extremely cheap, so it's not as heavy of a parameter model. and then once this finishes, what happens is the analysts or the main agent, the leader of the table, the war room, will go through and build a, final report for you of, uh, consensus.

But you don't have to stay within the screen. you can go back to your history. Y- you'll see here it's consolidating the results, and it's loading. But you can also see your history and what you have done. You can see presets. So if you design agents that you wanna keep, they'll stay there. and you'll get a finalized report when it's out. A nice PDF. You can download it, format it. It'll just be what the main agent's final result is based on hearing both sides' arguments and through the rounds and iterations.

um- I think it's real important for this because what we saw here is we're only at what? Not even seven credits yet, you already had both of the agents finished. But if you don't want it to pick what agent it uses, you can actually say, "Use, uh, Claude," or, or, "Use ChatGPT." Uh, and you can then select the reasoning levels, and you can select which clause. You could go with like, o3-mini, or you could go with a really heavyweight, you know, five point six, uh, Sol or, you know, the Astra 6, et cetera.

And same on the Claude side. You can say, "Use Fable." or you can say use, you know, one of the cheaper agents. So you have complete control, and again, complete transparency as to what's being done if you decide to have the, uh, main agent decide on your behalf. Awesome. And thank you, Glenn. This is the final PDF. You can export it, you can continue the conversation to have more rounds occur. like I was saying, go view your history, overview that

Credits, sectors, and the options screener

document in the future, And all that for less than ten credits, which, for, even the free account, you could run ten of these per month. you know, so that's pretty impressive, right? Not to mention all these tools you get access to free. lots of them for, for free, with your free account. And of course, you get much more capabilities with our different, paid tiers, which we're not pushing here. We'd just like you to try out some of these cool tools and let us know what you think.

Yeah. And o- on top of that, we recently added the sectors page. Just a quick snapshot of different sectors, and we will be adding in some heat maps here pretty soon. also the option screener, which is similar to the research terminal. You can see there's eight hundred thousand contracts that you can query and filter and sort by a plethora of options if you, wanna look at the valuation or which has the, largest theoretical edge according to Black-Scholes modeling.

And then you can come to the table and sort each column by any data points, save views, that kind of thing, or use our prebuilt queries that'll go through and run and show you the output. And we explain the strategy a little bit if you click Learn more, so

The community marketplace for skills and agents

you could understand. You said we're really focused on AI. Uh, just, you know, if we could look at marketplace, Connor, just so we can show people what's going on. So this is the community. Community is the GNG, anybody that's on GNG, you can go and upload your own version. But even more impressive is you can download all the ones that already are written, and it's completely open. So you get all the prompts, you get all, all of the instructions, everything, and then you can tweak it on your end.

So maybe if we're looking at an ETF look-through or the, or the rule of forty that I wrote, or a support and resistance zone mappers, which I also wrote, you, you can say, "Well, I don't really like the way Glenn did this. I just wanna tweak it a little bit." Great. You installed it yourself, and then you could upload maybe a new version with a, you know, some new twist that I wasn't thinking of, or whoever the author was.

So it's pretty amazing because, again, none of this is being charged for, for all this community's, add-ons that you get added for, right? That you have a large group of people that are contributing, which is pretty, pretty nice to have, right? Does a lot of the homework for you. Awesome. Thank you, Glenn, and that, also ties in really nice when it comes to agents, 'cause you can share both agents' skills and filter by different types up here at the top.

You can go through and say, "You know, I want a agent that's specific to screening," right? And it'll appear which would be mo- most relevant. So, and this feature set is, still growing. So we're listening to what people are saying, and we're continuing to add to it. The portfolio integrations and the workflows, together, the workflows primarily were the newest addition. and it's kinda like just having your own agent always watch for you. So it's, uh, similar to a lot of stuff I've seen with people trying to set up on their own who are more technical, like in, in cloud instances or, OpenCL, if that's still even around.

I saw people were trying to do similar stuff. But then, you know, it's hard to get all the data integrations and the data in place and have a community and different skill. And we, we have that all set

Data integrations and wrap-up

up, and it's, it's pretty unique. Yeah. 'Cause, I mean, it's Alpha Vantage, it's Edgar, it's Fin Hub. I mean, we have a lot of aggregate data, right, Connor? I mean, you've done an amazing job. You know, Connor is the CTO here, and he's done an amazing job of, you know, such rich data. It's hard to do that on your own. I mean, I'm technical. I mean, I had a computer science degree, and this would be, even with AI, a very difficult thing to do at this level of integration of all the different pieces together.

So- Even in the chat, when we used AI chat, I was counting, but I think we, we used the capabilities of, like, a few years ago, you would have to use fifteen different websites for the things that we went through for G Aerospace and the other stuff. You would at, at least fifteen. I mean, all the tools used to be, like, separate websites you would have to visit, start an account and so forth. And that, that has, I don't know, made me so much more efficient.

That's, really the fun thing. Yeah. A fun one that Connor's, uh, written on his own, he hasn't put it as a skill yet, and I hope he does, is he can actually pull out all the screenshots within an SEC filing or, you know, their earnings call, you know, where they do the presentation. Of course, you can get the PowerPoint, but you can have it in context with what they're talking about, right? And so then, especially for people writing, the author types, which are, you know, we're not nearly as popular as the people that are gonna be wanting to use all this, but it's still very useful information when you start seeing all this put together in context.

It's the important thing. Yes, exactly. Awesome. And yeah, Glenn, that's something I'll be, I'll be getting in here soon. Um- It's on the roadmap.

About this video

Connor (GNG Research CTO), Glenn, and Leo walk through the GNG AI Console, also called the GNG Analyst, and show how they use it in their own research.

What we cover:

• Searching SEC filings by meaning, with cited sources

• Workflows that run research on a schedule or when a market event happens

• Skills and agents, with the Rule of 40 as an example

• Memory and file uploads that give the AI context about how you work

• Backtesting and Monte Carlo forecasting run on the platform's own engines

• Checking an EPS outlook and how many analyst estimates it is based on

• Advanced charting with crowdsourced estimates, plus premium and discount checks

• Market sentiment, insider trades, and ETF look-through

• The War Room, where AI agents argue a question and a lead agent writes a final report

• The options screener, the sectors page, and the community marketplace for skills and agents

The tools shown support your own research. Nothing in this video is a recommendation to buy or sell any security.

For education only. Not financial advice.

gngresearch.com

X: https://x.com/GNG_Research

LinkedIn: https://www.linkedin.com/company/gng-research/