- Speaker #0
Hey everyone, super stoked for this deep dive. We're going to check out Qolaig.
- Speaker #1
Qolaig?
- Speaker #0
Yeah, they're a French company doing some really cool stuff with AI for productivity. Remember how you were asking about how AI can help, like, actually boost how much you can get done?
- Speaker #1
Oh yeah, definitely. I've been trying to wrap my head around all that.
- Speaker #0
Well, it seems like Kolag is right at the cutting edge of all that. Yeah. We've got slides from one of their presentations, so we're going deep on those. Should give us a good look at how they do things.
- Speaker #1
Sounds good.
- Speaker #0
All right, so first things first. Colag is all about making AI that's tailored for you. No off-the-shelf stuff here.
- Speaker #1
Oh, so like custom-made solutions, right?
- Speaker #0
Exactly. They find those little kinks in your workflow, you know, the stuff that slows you down, and make an AI tool that fits right in.
- Speaker #1
Interesting. So it's not like they just have, like, a set list of products or something. No,
- Speaker #0
totally bespoke. And they've got a really clear process for how they do it. Three pillars, they call them. Orchestration of tools and AI, UX design. And then data protection.
- Speaker #1
Ah, so they're not just slapping code together. They're thinking about how people will actually use it.
- Speaker #0
Yeah. The UX design part, make sure it's all user friendly, you know. And of course, got to keep that data safe. That's the data protection pillar.
- Speaker #1
Makes sense. I'm seeing a lot of companies talking about user-centric AI, but it seems like Colog is actually walking the walk.
- Speaker #0
Totally. But the thing that really stands out to me is their approach. They call it pragmatic.
- Speaker #1
Pragmatic.
- Speaker #0
Yeah. It starts with this AI workshop, really getting to know what the client needs.
- Speaker #1
So like figuring out the problem before jumping to solutions.
- Speaker #0
Exactly. Then they prototype stuff, develop it, and finally roll out the finished solution, all tailor-made and scalable.
- Speaker #1
Scalable, good point. So it can grow with the company, but it doesn't stop there, right? They also keep things running smoothly over time.
- Speaker #0
You got it. Ongoing maintenance and making sure the solution evolves as things change. It's a real partnership, not just a one-time thing.
- Speaker #1
That's impressive. So many companies just kind of drop the product and run.
- Speaker #0
Not cool. Like it seems like they're in it for the long haul. And let's talk about LLMs, large language models. They really know their stuff.
- Speaker #1
LLMs, right. Everyone's talking about those.
- Speaker #0
They work with everything. Open source models like Mistral AI, Lama 2, even the big ones, Google Gemini and Azure Open AI.
- Speaker #1
Wow. So they're not tied to just one platform or model. That's pretty flexible.
- Speaker #0
Super flexible. They can pick the right tool for the job, considering data privacy, all that good stuff.
- Speaker #1
I bet they've run into some interesting challenges trying to get all those different systems to play nice.
- Speaker #0
Right. It's got to be tough. Sadly, the slides don't really dive into their challenges.
- Speaker #1
Maybe you can ask them about that later.
- Speaker #0
Good idea. But hey, their client list speaks for itself. Action Lodgement, Folk, Data Scientist, Activa Capital, Elcano Asset Management, just to name a few. Whoa.
- Speaker #1
That's a diverse group, so they're not just catering to one specific industry.
- Speaker #0
Definitely not. It shows they can handle all sorts of challenges, all those different industries.
- Speaker #1
It's making me more curious about how their AI actually works in practice.
- Speaker #0
Well, lucky for you, we're about to see the magic in action. First up, Colag's Copilot Database solution.
- Speaker #1
Copilot Database, huh? Sounds intriguing.
- Speaker #0
They built this for a big real estate company. Imagine being an executive there. You need data to make decisions. but it's locked away in a database. Ah,
- Speaker #1
yeah, so they have to go through technical teams to get the data, right? Like sending SQL queries and waiting for the result.
- Speaker #0
Exactly. Tons of wasted time waiting for someone else to get the info you need. Ugh,
- Speaker #1
I can feel the frustration already.
- Speaker #0
Right. So what did Qolaig do? They made a chatbot.
- Speaker #1
A chatbot? For a database? Yeah,
- Speaker #0
but not just any chatbot. This one lets the execs talk to the database directly, using plain English.
- Speaker #1
Whoa, hold on. So they don't need to know SQL or anything, they just… type in what they want.
- Speaker #0
Exactly. It's like having a personal data analyst right there in the chat. It's seriously impressive. They connected to Snowflake's API, used Azure OpenAI and Langchain for prompt engineering, and put it all together in a super sleek interface.
- Speaker #1
That's a lot of moving parts.
- Speaker #0
I know, right? But the result is this seamless, intuitive experience. No more waiting around, just instant access to the data they need.
- Speaker #1
It's amazing how this kind of tech is changing how companies make decisions. They're not stuck waiting for someone else. They're empowered to get what they need when they need it.
- Speaker #0
Totally. It's not just about efficiency. It's about putting the power back in the hands of the decision makers.
- Speaker #1
I love that. It feels like Qolaig is really getting that.
- Speaker #0
Right. This co-pilot database is just the beginning. I can't wait to see what other cool stuff they've come up with.
- Speaker #1
Me neither. Let's keep going.
- Speaker #0
All right. Next up is something I think we can all relate to. The dreaded Review to Press.
- Speaker #1
Review to Press. Is that like a press review?
- Speaker #0
Yeah. You know, having to read a ton of articles every day just to keep up with what's going on.
- Speaker #1
Oh, yeah. That can be a real time suck.
- Speaker #0
Big time. So imagine you're an analyst at an investment fund, spending like two hours every morning putting together a press review.
- Speaker #1
Two hours? Ouch.
- Speaker #0
Right. So Qolaig decided to automate that whole process, free up those analysts'time. They made a tool that automatically creates these press reviews, and you can customize it for different layouts and how often you need it.
- Speaker #1
That's super clever. But how does it even know which articles to include? Ah,
- Speaker #0
good question. This is where those LLMs come in again. They use web scraping to pull in a bunch of articles, and then the LLMs figure out which ones are actually relevant based on the client's criteria.
- Speaker #1
Ah, so it's not just grabbing anything with a keyword. It's actually understanding what the client cares about. Yep.
- Speaker #0
And to make sure it doesn't miss anything important, there's also a human verification step before the final press review is ready.
- Speaker #1
So it's a mix of AI efficiency and human oversight.
- Speaker #0
Exactly. They're finding that sweet spot. This review to press solution takes something that used to be... a total chore that makes it like effortless.
- Speaker #1
Totally. And it frees up those analysts to focus on the stuff that really matters, the high level analysis and strategy.
- Speaker #0
You're getting it. I'm noticing a pattern here. Qolaig finds those tasks that just suck the life out of you and finds ways to automate them, even make them better.
- Speaker #1
And it's not just about getting rid of the boring stuff. It's about improving the whole workflow.
- Speaker #0
You got it. And you know what? They do the same thing for comparing bids from different providers.
- Speaker #1
Oh, bid comparison. Yeah. Yeah. Yeah. I can see how that could get messy, especially when you have a lot of providers for complex projects.
- Speaker #0
Imagine being a real estate developer, right? You've got all these bids to go through and it could easily take half a day per project just to process and compare them all.
- Speaker #1
Half a day. Jeez, that's a lot of wasted time.
- Speaker #0
Major times suck. So Koleik stepped in and built this AI powered solution that streamlines everything.
- Speaker #1
So the AI compares the bids for them?
- Speaker #0
It's even cooler than that. Providers submit their bids through this custom interface. Then a trained LLM swoops in, processes, and aggregates all the data. Okay,
- Speaker #1
I'm following so far.
- Speaker #0
Then the AI creates this awesome DAC board that summarizes everything in a way that's super easy to understand. The developer can compare bids side by side and make informed decisions no problem.
- Speaker #1
That sounds amazing. It's not just about saving time. It's about reducing errors, being more accurate, and ultimately making better choices.
- Speaker #0
100%. And the best part is, Coal Ag tailored this whole solution specifically for real estate developers. It's not some generic tool. It's built with their needs in mind.
- Speaker #1
That's smart. It makes it so much more effective.
- Speaker #0
I'm telling you, they really have a knack for understanding the specific challenges businesses face and using AI to make things better.
- Speaker #1
I'm convinced. So what else have they got up their sleeve?
- Speaker #0
Well, next up is an area where AI can really shine. Customer support.
- Speaker #1
It's interesting. I've seen some pretty basic chatbots out there, but I'm curious to see what Kolag has done.
- Speaker #0
Me too. But first, let's take a quick pause. We'll be back in a jiffy. We're back, ready to jump back into the world of Kolag and how they're using AI to level up customer support.
- Speaker #1
Definitely. We were talking about how they really understand that great customer support is about more than just answering questions quickly.
- Speaker #0
Exactly. It's about making sure people can find what they need easily. Remember that feeling of trying to find something specific in a huge document?
- Speaker #1
Ugh, yeah, like trying to find a needle in a haystack. You know it's in there somewhere, but...
- Speaker #0
Impossible to find. Coal Lake tackled that exact problem with one of their clients, a big social housing provider.
- Speaker #1
Social housing? Interesting. What was the issue there?
- Speaker #0
Think about being a tenant, right? You need info about your lease, or maybe you have a maintenance request.
- Speaker #1
And the answer is probably buried somewhere in pages and pages of technical documents.
- Speaker #0
Exactly. So, what happens? The tenants get frustrated and the support team gets flooded with calls and emails.
- Speaker #1
Yeah, no one wins in that situation.
- Speaker #0
So Kolag came in with her AI magic. They built a chatbot.
- Speaker #1
A chatbot, huh? To handle tenant inquiries?
- Speaker #0
But not just any chatbot. This one can actually understand and use all that technical documentation. Wait,
- Speaker #1
so the chatbot can read and understand all those documents?
- Speaker #0
It can. Tenants can ask their questions in normal everyday language, and the chatbot finds the answers from the documents.
- Speaker #1
Wow, that's pretty amazing. No more sifting through pages of jargon. Just ask and you shall receive.
- Speaker #0
Exactly. It's like having a 247 support agent who knows everything about the company's policies and procedures.
- Speaker #1
It's proactive, too. It's solving a problem before it even becomes a huge headache for everyone.
- Speaker #0
Right. It's all about giving people the information they need when they need it.
- Speaker #1
So it's a better experience for the tenants and a more efficient support team. What's not to love?
- Speaker #0
Exactly. Qolaig is all about finding those practical applications of AI. And you know what else they're good at? Streamlining reporting processes.
- Speaker #1
Oh, reporting. Don't even get me started. I've lost so many hours to spreadsheets and data entry.
- Speaker #0
It's a pain, right? And it's so easy to make mistakes when you're doing everything manually.
- Speaker #1
Tell me about it. And for businesses, those mistakes can have serious consequences.
- Speaker #0
Kulag knows this, and they've built some really slick AI solutions that automate those reporting tasks. They've even done some impressive work with companies in the healthcare industry.
- Speaker #1
Healthcare reporting. That sounds intense. What kind of challenges were they facing?
- Speaker #0
They were working with a group of medical centers that were struggling to keep up with their key performance indicators. Their old reporting system was slow, clunky, and prone to errors.
- Speaker #1
Sounds like a recipe for disaster. So how did Kulag help them out?
- Speaker #0
They built a system that... It automatically pulls data from all these different sources, processes it using a generative AI model, and presents it all in a nice, clean dashboard.
- Speaker #1
A dashboard. Using something like Power BI, I bet.
- Speaker #0
You got it. So instead of spending days wrestling with spreadsheets, these medical centers can now see real-time insights with just a few clicks.
- Speaker #1
That's huge. It's not just about making things easier. It's about giving them the tools they need to make data-driven decisions quickly.
- Speaker #0
Exactly. They can spot trends. catch potential issues before they blow up, and adapt their strategies much faster.
- Speaker #1
It's like they've gone from driving a clunky old car to having a sleek sports car.
- Speaker #0
Perfect analogy. And speaking of fast-paced environments, let's shift gears and talk about finance.
- Speaker #1
Finance. That's a world where time is definitely money.
- Speaker #0
No kidding. Coleg has been making some serious waves with their AI solutions for investment banks, particularly in the world of mergers and acquisitions.
- Speaker #1
M&A, huh? I can only imagine the amount of paperwork involved in those deals.
- Speaker #0
Mountains of it. Think financial statements, contracts, market data, all that fun stuff. It's crucial to do your due diligence, but it takes forever.
- Speaker #1
So how did Colag tackle this beast? Did they create an army of AI lawyers or something?
- Speaker #0
Not quite, but close. They developed a tool called the CoPilot M&A. Think of it like an AI analyst that helps bankers navigate all that complexity.
- Speaker #1
So instead of drowning in documents, the bankers have this AI assistant to help them make sense of it all.
- Speaker #0
Exactly. They trained this co-pilot on a huge data set of historical deal data so it knows its stuff.
- Speaker #1
That's smart. So it can answer questions, spot potential red flags, and offer insights based on past deals.
- Speaker #0
You got it. It can do all that much faster and more accurately than a human could.
- Speaker #1
This could be a game changer for the M&A process. Not just about saving time, but about making smarter decisions and avoiding costly mistakes.
- Speaker #0
Absolutely. And it shows that Colug isn't afraid to tackle those complex high stakes challenges. They're really pushing the boundaries of what AI can do.
- Speaker #1
I'm seriously impressed. They're not just building cool tech, they're actually solving real problems and making a difference.
- Speaker #0
Well said. We've covered a lot of ground today, from real estate and finance to healthcare. But there's one more case study we need to discuss, and it takes us back to customer support.
- Speaker #1
Oh, another customer support solution. What have they come up with this time?
- Speaker #0
This one is all about personalization. Coleg worked with Data Scientist, a big edtech company in France, to create a customer support experience that was tailor-made for their users.
- Speaker #1
Personalized customer support. I'm all ears.
- Speaker #0
Me too. Let's dive in. We're back and ready to wrap up our deep dive with one last look at Coleg's customer support solutions.
- Speaker #1
Yeah, you were saying something about personalization before the break. I'm really curious to see how they're using AI to make that happen.
- Speaker #0
So this case study features Data Scientist. pretty big ed tech platform over in France. They were growing super fast and their customer support just couldn't keep up.
- Speaker #1
The classic scaling problem happens to the best of them.
- Speaker #0
Right. They needed something that could handle all those questions, but also keep the quality high.
- Speaker #1
So instead of just throwing more bodies or a basic chat bot at the problem, they went with Colog.
- Speaker #0
Exactly. Colog went all in on understanding their users. They dove deep into data scientists'customer service history.
- Speaker #1
So they basically built a knowledge base of all the common questions and answers, right?
- Speaker #0
Exactly. But they didn't stop there. They took it a step further and trained an LLM on all that data.
- Speaker #1
Ah, so this is where the AI magic comes in. But why not just use a regular chatbot?
- Speaker #0
Because Colag gets that context matters. They wanted to create a co-pilot that was specifically tuned to data scientists, platform, and users.
- Speaker #1
So it's not just spitting out generic answers. It actually understands the nuances of data scientists.
- Speaker #0
Exactly. It's all about providing relevant and helpful support, not just quick answers.
- Speaker #1
That's really smart. But how did they make sure the co-pilot could actually understand what users were asking?
- Speaker #0
They used a technique called vectorization with elastic search. Basically, it's like translating human language into something AI can understand.
- Speaker #1
Vectorization, huh? I've heard that term, but never really understood what it meant.
- Speaker #0
Think of it like this. Each word and phrase gets turned into a set of coordinates, kind of like on a map.
- Speaker #1
Okay. That makes sense. So the AI can see how different concepts are related to each other.
- Speaker #0
Exactly. It goes way beyond just matching keywords. It can actually grasp the meaning behind a question, even if it's phrased differently.
- Speaker #1
Wow, that's seriously cool. So did this super smart co-pilot actually work?
- Speaker #0
It did. Data scientists saw a huge drop in response times and much happier users.
- Speaker #1
Makes sense. When you get the help you need quickly, you're going to be happy.
- Speaker #0
Right. And the best part is the support team had more time to focus on those really tricky cases that the co-pilot couldn't handle.
- Speaker #1
It's a win-win. It really shows how powerful personalized customer support can be.
- Speaker #0
And that brings us to the end of our deep dive into Colog.
- Speaker #1
Wow. What a journey. They're doing some seriously impressive stuff with AI.
- Speaker #0
Right. They're finding creative solutions to those everyday problems that can really slow businesses down.
- Speaker #1
And they're not just focused on the tech itself. They're focused on how it can actually help people.
- Speaker #0
Couldn't have said it better myself. So, dear listeners, we challenge you. How could AI change the way you work or even live your life? What could be made easier, faster, or more efficient?
- Speaker #1
The possibilities are endless. CoalLag has shown us that with a little creativity and a focus on human needs, AI can truly make a difference.
- Speaker #0
Thanks for joining us on this deep dive. Until next time, keep exploring and keep innovating.
![[Episode Generated by AI] Deep dive into Qolaig cover](https://image.ausha.co/TNVBrhtJu7JnO9gEbc1X6zgrTW2nPfo30CifD8lw_400x400.jpeg)
