Allie Systems Pitches at Startup Battlefield | TechCrunch Disrupt 2023

Okay next one from Mexico City Mexico we Have La systems presenting for the Company is Alex sandal give him a round Of applause and bring them on Out hi I'm Alex CEO of Aly systems we Make machines speak and translate them Into actions on the factory Floor Slide Please Manufacturing Systems are broken Strained Supply chains inefficiencies And downtime we've thrown robots and Automation at the problem but downtime Is still costing industrial players over 11% of their yearly Turnover this is a 1.5 trillion doll Sinkhole it's urgent that we fix the Foundation of the real World economy now The time is now because Supply chains Are rapidly Changing nearshoring is becoming the Strategic pivot moving manufacturing Closer to its Consumer we learned the hard way that The world cannot stop running because of Its dependency on one country a war Cannot put our food supply chains at Risk it's imperative that we move Manufacturing into new markets into a More stable order of Supply Chains even as even as Mexico and Latin America fa Tailwinds the supply to meet Demand is going to be an uphill battle

Take monter one of the states where we Have an office this state alone produces $28 billion in contribution to Mexico's GDP next year they have announced that They will host a Tesla gigafactory Foxcon and Tata so how are they going to Keep up with this demand we're talking About an $800 billion tsunami of Production demand that is coming their Way and they need to get ready now this Is why we've built Ali Ali is a capex Light infrastructure to turn Legacy Operations into intelligent Factories one that keeps learning with Every Action we started with this device right Here we call it Ali's Gateway it has Built-in code that supports the most Widely used connectivity protocols Allowing us to conent to over 90% of Industrial Machinery categories sensors And Plc's back to Presentation this is allowing us to Connect to the factory and pull in data From processes from machines And from machine health we're collecting Over 200,000 data points per line per Facility this is an untapped data gold Mine until now with built-in Erp Integration and maintainance data this Device becomes Ali's Central source of Truth this is how we're installing it on The client site and it's pulling in data

Points from cost maintainance machine Health productivity And building the brain of the Factory this is what enables our Software move to Demo one of our cement client loses $500,000 every hour their oven is not in Production mode we're tracking every Single one of their Machines machine learning models detect When there's irregular activity or Machine Failure but here's what makes makes us Different to any other manufacturing Software out there we're not a Monitoring solution we're an intelligent Being we tell factories what's the Probable cause and what's the Recommended course of action years and Years of factory knoow now live on in a System that makes predictions and makes Recommendations in our efficiency module We're tracking Productivity quality costs and Availability If you have a mixture d a mixture that The density is not hitting in Target We'll let you know is there a product Wastage in one of the stations will tell You why where it is and what to do about It the beauty of our model is every Single action improves our model's Algorithms creating Network effects that Ultimately benefit every single customer

On our Platform but we thought really hard how In an industry where every single second Counts how can we make decisions even Faster and simpler so we built a Chatbot we call it the Ali AI assistant But this assistant is trained with all Of your productivity data so we can ask It anything about your operation Ernie Let's ask it to please graph operational Efficiency by station and date in the Last month We've asked Ali things like please tell Me the root cause of all my downtime to Take to make Comparisons and it's built in with Generative bi so it gives you Visualizations and these are problems That would have taken two weeks to solve Weeks of data crunching a squad of Engineers and now you can solve it in Second there you Go so let's back to presentation Let's imagine a world with alley enabled Factories running AI powered operations To make factories respond quickly to Market demands making manufacturing more Efficient and more sustainable we're Very excited to say that in 12 months we Have achieved seven figures of Revenue Averaging productivity increase of 15% Over $400 million are trained and Monitored through Ali so if you believe In

In our mission of stabilizing Supply Chains in in a world of more efficient Manufacturing come join us and let's Build smart operations together thank You we'll go to Rebecca yeah so uh so I Actually spent some time in plants at One point in time in Mexico I helped Launch Proctor and gambles as a couple Divisions in Mexico so would love to Know about um your go to market so Really impressive $400 million or Transactions going through the system Tell us about that tell us who what You're doing you know those uh prefa Concepts or those contracts yeah so our Go to market starts from building uh an Industry case study from the one of the Top three manufacturers in the region We've built a network of Angel Investors Of key industry players in the top Manufacturing cities in the north of Mexico and in key industrial cities in Latin America angels from Mexico from Mexico and Latin America they're Introducing us to the owners and the Family offices of the top industrial Conglomerates five to from billion Production facilities onwards and and That's how we uh Target and build a case Study and then that we have a trickle Down effect to all the other players in The segment is Slim an Investor so soon hopefully Okay Jacob yeah just to kind of double

Click on that so what is the timeline And the progression look like from like You get one of these angels on the c Table hopefully they start connect you With these folks do a proof of concept When does that turn into a mature Customer and then does that continue to Scale or is that kind of study state Revenue with that plan the the beauty of Our model is because we start from the Top it's a CEO mandate um and we can Move very quickly we've closed deals as As quickly as 1.5 months um we are Transforming the operation of the Facility and because it becomes from the COO office we work with all the Departments like quality maintainance uh To transform uh the operation and that In in total from signing contract to Having a a live s side it takes about Three to 3.5 months what's the just kind Of a double click on that what is kind Of The NPS on the people that are maybe Not a part of the top down decision when This thing gets dropped into their plant Like are they happy to have this thing Or is this like hey we got told to do This so right now that a very Interesting question we've we are Realigning the incentive model on all Levels so that we not only make their Life easier because it's really easy and Simple to use our platform but if we Have efficiency increase those

Uh gains also get translated to the Operator okay great yeah go ahead so This is kind of observability for Manufacturing right and so how similar Are those plans so if you if you make One plan successful and you realize Certain gains how easy it is to go from Like car manufacturing to like medical Device manufacturing are they going to Be similar or like wildly different Right I think our models are um really Effective a in a at a segment level so Steel manufacturing is a big segment for Us food and beverage and we find about 80% of similarities between the types of Machineries and processes that are being Used we've invested a lot in Configurability so starting with Ali you Can configure your notifications you can Configure your machines you can create Users allowing us to have a very Scalable scalable platform from day one And then um when is the magic moment so Let's say you shoot hands with a CEO Start deploying and when is the first Time you can demonstrate value for for Your solution we we've had a lot of uh Aha moments I think the first one is When they start seeing patterns like all Of this data lives in silos they have a Silo for maintainance a silo for Efficiency and they've never had a Connected source of data that's finding Patterns and correlations between the

Data so when you tell them hey one of This processes and this component is Costing you 5% of your operational Efficiency and you can get that through Ali's assistant that's an aha moment for A supervisor and for an owner and Finally um oh go ahead uh so you Demonstrated the chatbot and and I'm Trying to parse in my head is it just Because AI is hot or you see that Chatbot to be the primary interaction With the system the the concern I have Is like models tend to hallucinate and So when you ask questions and in playing English how do you ensure that that the Reports are accurate because like There's a Machinery on the back end that You know interfaces with the data Warehouse and whatnot that that like Generates those report really good Question we we didn't b a chatbot Because of a trend we built it because They has task forces that need to do Analysis and they take weeks to do it so We really thought how how do we make Decisions really really easy and really Really simple we have a data pipeline That cleanses all of the data from all The reports so we have models that do Unsupervised uh machine learning to do Classification and then we have have Semi-supervised and supervised models That are basically cleansing the data Goes through a pipeline and then when

It's classified it goes into a database That this queries to give them effective And accurate information thank you Guru As the underlying machines in these Manufacturing sites get smarted Themselves and they're alerting the Users and also talking about sort of Intelligent uh manufacturing do you find Yourself at odds competing for the same Dollars cuz they're trying to sell Monitoring for their machines and so are You for you uh interesting question as Well we found that automation does not Equal digitalization so you can have a Very very smart seens machines from 2023 But that machine will and will have a Screen and it's monitoring everything But it's completely disconnected from Everything else you don't know what a Failure in that machine is causing Quality what it's causing your Operational efficiency because it's Living by itself so we are agnostic we Connect to all types of Brands old and New and we're able to do um correlations Between those data states which a which A manufacturer can't can't do at this Point what is the diversity of equipment Manufacturers in a typical plant is it All seens or is very very very large Like the max talking we're talking 250 Brands in in one facility okay all right So they they would have to purchase 250 Software from 250 Brands right we're out

Of time here thank you so much give Alex A round of applause thank you so much Thank you