AI for Supply Chain

[@jayshooooo]

AI, blockchain, digital twins, and IOT will revolutionize the way we plan, source, make, deliver and return. I love your channel, thank you for all this mind blowing content.. I just wish we can see your results of the analysis after the code. :)

[@jonashubotter6686]

"There are all sorts of middleman in the middle." 😁👌

[@sim021ful]

Jonas Hübotter this nigga

[@wolfisraging]

Great day, great topic, great video. I mean it's insane

[@SirajRaval]

thanks rishik

[@stevenlong5431]

Hi Siraj, I am supply chain professional and just starting to learn AI. Thanks for your how to learn AI in 3 months video, I am re-studying linear algebra and calculus now, and learning a little bit python step by step. A bit slow, you know work, family daily life etc... I am truly believe Supply chain is a great place to apply AI. And there is a loooot of money in this industry(supply chain management industry). It’s probably 10 time bigger than self driving cars. And believe me I’m being conservative. Compare to AI, those Current ERP giant company such as SAP and Oracle are just some dying beasts. The key to supply chain world is purchasing orders. A sample PO will change everything, because that’s everything begins. Forget about save 5% on transportation cost, that’s small money compares with reducing 5-10% inventory, 5-10% demand forecasting accuracy. I knew a company last year only, they spent 13 millions dollars and increased 2% demand forecasts accuracy and they call that was a very successful project. If you need some information about developing AI in supply chain, you can contact me anytime. So, my goal is by the end of this year or early next year I wish I can develop something to help people create purchasing orders. Hopefully, I can achieve it. Again, thanks for your video.

[@MrAbdulawal]

I liked your comment for almost similar situation. As a supply chain professional, I believe AI shall transform Supply Chain and business as a whole big time. Learning..

[@Moriahvivatv]

Can I get your Contact?

[@muhammadsami7711]

Love your videos. I have a suggestion. Man, It would be awesome if you give the intuitive concept of the NN's in a particular amount of time and ALSO write the code(from scratch not using a high -end library like KERAS). Because Code is the best part. It also tells us how to make the model and get the intuitive sense of making it.

[@joliver1981]

I agree with your point that writing a model from scratch would give a better intuitive sense. In fact I have actually written a NN from the ground up using no libraries, just pure Python, (I even used for loops in place of matrices) and I actually did end up with a better understanding of the mechanics of a NN. Of course using for loops is absurd but my goal was simply to gain a true understanding of what is happening during training. NN's are like onions where you have to peel back the layers and there are many levels or degrees of understanding.

[@muhammadsami7711]

totally, can you please send the code you are talking about to [e-mail] I want to take a look at it.

[@joliver1981]

Sure thing. I'll send it over tomorrow when I get home to my desktop PC.

[@mickey5370]

James Oliver dude, I’d love to take a peek too, please?! [e-mail]

[@muhammadsami7711]

I appreciate that : )

[@bulvinescatologist2374]

I do see sum of what you say about "AI" but do not see how this can help us in Alaska yet. Thank You Sirej Raval .

[@X_platform]

Your sidetracked game is so strong lol Always go to tangent topics

[@SirajRaval]

haha. i try to keep it interesting

[@EmissaryOfGorz]

@@SirajRaval It’s good on my end! Haha I love the enthusiasm! Keep it up!

[@le7724]

thank you so much for this video, i want to build a start up and i guess you just gave me so many ideas ! thank you so much !

[@SiilverEdit]

did you

[@le7724]

@@SiilverEdit Hey, its been so long since i watched this video haha, i didnt make something with AI. but did build a small web agency 😅. WIshing you luck with your peojects too :)

[@SiilverEdit]

@@le7724 can i check your work?

[@woolfel]

Lol, IBM is probably the worst company to talk about supply chain. Watson also sucks compared to the other players. I worked at IBM and still have friends working at IBM. The problem is their software is expensive and out of date.

[@evolutionisnear]

Would you be able to recommend a different example? IMO still better than manual entry in SAP or excel

[@woolfel]

Honestly every software built for supply chain has manual entry. Unless someone invents a machine that can read people's thoughts, some manual entry is going to be required. The problem with IBM isn't the manual entry part. It's the fact that many workloads today can't run on IBM's software. In fact, for many years IBM has been telling their customers to export data from their BI tools into Hadoop and use hadoop to do analysis. Add on top of that the horrendous pile of poorly organized documentation, and IBM consultants that don't even know their own products. The top management at IBM are lost. There's still good engineers at IBM, but bad management kind of ruins it all. On top of that, IBM hasn't been able to attract or keep AI talent. Not that SAP is any better. Microsoft on the other hand is doing some great work with BrainWave.

[@evolutionisnear]

You make a very good point in terms of manual entry, the question is how much, how often and those small opportunities to shave off minutes without a reduction in safety or quality (like some companies manually entering OEE sheets, or manually entering truck load weight). RFID and GPS auto-input are by now industry standards, but there are still small manufacturers and transportation companies that fail to incorporate which is a huge opportunity to streamline, not only for efficiency but quality control which consumes a massive amounts of manpower (for good reason!). But I REALLY appreciate you describing the flaws of IBM, sounds like a radical HR department IMO, and if such a well-known company has such a deafness to feedback... it goes to show the potential of benchmarking when it comes to semi-intangiables (consultant proficiency, ease of software transitioning, management constraints for example). The book Billion Dollar Lessons underlines how destructive management can be when they can't step back and evaluate where they are not needed or unequipped to deal with the issues at hand. Again, thank's for the description! (For now, it's rabbit hole time with BrainWave)

[@woolfel]

Having RFID and/or QR Codes help reduce manual entry. If you look at the recent video about phone QA process on linus tech tips, you'll see the more advanced companies use QR codes to improve supply chain and defect tracking. Aside from mainframes, IBM got out of the PC business over a decade back. They may understand supply chain at a high level, but that's different than real world experience. If we're talking about supply chain, the two companies that suffer the most are Boeing and Airbus. 100% of the delays for both companies are due to supply chain issues.

[@evolutionisnear]

I couldn't find the link you were referring to, but the entire NPU and DNN FPGA was a theme of that rabbit holing that would be interesting in how it links to easing QR code detection and/or quality control as materials are processed through the supply chain, even in terms of adequate course trajectory for the two examples, especially to reduce the small delay and cost of having a 'transparent-real time supply system'. Hmmmmm!

[@DevOpsBhang]

today for me and it is 6 years ago, wow amazing

[@0rajatvashistha449]

It is the last nighy only when I am planning stuff, for dynamic pricing using ML and today you comes up with this video... Hahahahaha

[@randombrandon12]

Please post more videos on supply chain and a.i. This is an extremely important topic.

[@ehabhamdy2012]

I am working in the supply chain department, we have similar data that you showed in the video but much uglier. I spend a lot of time normalizing this data for the purpose of documentation to inform sales department to pay freight costs. I am just getting started in the company and I am so worried if that is not normal about the very ugly data we receive from different sources (carriers and the company plants). Thanks for the video, I will definitely work on a project to extract some insights from this data.

[@jeremycornish-ford2435]

Siraj be hittin licks.

[@Go4Ceo]

I am quite possibly the least tech person in my 40s but this and much of what this world is has recently been interesting me. I’m an #Entrepreneur by nature and very excited to have come across this channel in my search of AI & Supply Chain. Thank You Siraj

[@poushalybd]

It is insane! I have a supply chain background, from the army, had done all these things manually. Now I feel the insane loss, of not having done it the right way! insane, insane, insane to the nth level!

[@hemanshuchotaliya1182]

Dear Siraj, that was amazing video. I am studying right now "Technical Procurement engineering" in germany. So during study I got one paper to write it called "Artificial intelligence in supply chain management"., so that this video I found very useful for my work. But right now I am still writing this paper. Can you share me recent infos regarding this topic, this willl be very helpful for me.

[@Suro_One]

Hey Siraj, thanks for always listening to messages. This was a problem I was facing and you answered it clearly! Kudos!

[@shamaldesilva9533]

Yes !!!! i really needed this for my university project =)

[@eddisonlewis8099]

Siraj, the video on use of Artificial Intelligence in Supply Chain is totally awesome

[@chrisbeasty]

Ahahaha I lost it at: ' You guys get money, I get happy that's just how it works...' ... .. 'and i got my paid course and book sales and my ad revenue my patreons AND SPEAKING DEALS' 😂😂😂

[@shivams417]

Awesome initiative bro, really its like working on live project for real world usecases. Thanks a ton.

[@ramsunil4317]

thanks siraj . me being manufacturing engineer ready to transit to Data scientist after watching this video.

[@sherkhan_5050]

This is what I was looking for. Thanks a lot 👍😃

[@rickfelix2000]

In the same way that you could predict the cost to ship based on the data set, could you apply to a different problem where you want to know when is the right time to bring a product to market? Sometimes the success of great ideas is all about timing.

[@dhananjay9549]

i am new in ai field what about accuracy of cost prediction ? what we do if our data-set size is small for predicting cost ?

[@ScorchedEarthRevenge]

You still need a hold-out test set to determine test accuracy. But for training and cross validation you can use n-fold cross validation (usually with 10 folds). Also make sure your test set has a representative sample of positive and negative observations.

[@amishdoshi9453]

This is dope! One question to the community, what are the best sources to learn about upcoming startups in the field of AI.

[@wadanda]

Awesome vid Siraj!! Lots of pointers to dig into, thanks brav !

[@musficamurshidopshori2512]

Hey! Your videos are really cool. Please make more videos on Machine learning and artificial intelligence. If possible, then also on thesis related to ML.

[@PP-rj4gv]

Siraj, It was amazing video as usual. Suggestion, Please do the real-time use cases from A to Z. Like, pre-processing to saving the model and how to retest the saved model with another dataset. If your time permits. Thank you in advance for listening. You have a great personality which is god's gift and keep it up.

[@arcade-fighter]

Engineers working on Supply Chain are very tough in maths and statistics with its deterministic methods such as simplex. so it is hard to chagen their mind into artificial intelligence but we will try wih your advices dear Siraj. You have done a good job as usual explaining these bussines oportunities a little compressed but we also can stop the video :). Please can you talk about Machine Learning applied to Geographical Information Systems?

[@souradeepmajumdar4167]

Hi Siraj, Big fan following your 3 months machine learning study plan presntly. Can you please explore AI in chattered accountant and civil engineering or mechanical egineering designing aspects. Or atleast provide the thought process in this disciplines. super excited to hear your thoughts on the same

[@mohamedabbes6064]

This is extremely helpful for me as an IE thank you very much Siraj

[@YattishRamhorry]

Hi Siraj, Great video! Could you maybe do a video for the Postal & Parcel delivery service as well? I think that space could be ripe for disruption at the mo as well

[@alemazzuca]

Thanks for the video, very clear. I have one question, can I do the same but with sklearn only? not with Keras. Maybe with LinearRegression or SVM or Knn?

[@ajinkyaraut2533]

hi Siraj, I have a question regarding the logistics firm. I have been given the sales data and the costing of the vehicles of a logistics firm. So currently the firm has a 1.5 million USD turnover and there's a new client who,s business would be of worth 1.5 million USD alone. So how should I expand my business to the new client with the current resources available and which all resources to buy and all such detail stuff to make a clear idea.

[@arjunsinghyadav4273]

Great Vedio Siraj, First feedback so I would stick to the bottom line, Can you please make a topic on how to do CONTROL and MEASUREMENT USING AI. Meaning a tool for companies to build strategy in targeting a specific area that is most suitable and then be able to measure the impact of the choice continuously.

[@holisolsupplychainmanageme9202]

AI is intelligence exhibited by machines, or when machines mimic or can replace intelligent human behaviour, such as problem-solving or learning. The beauty of AI-based solutions is that they learn and drive continuous improvement over time. They get more precise and sophisticated as they gather more data and more experience. The sooner you start, the better the results you’ll see in future, and the further ahead you will be. With the right AI solution in place, you can outpace your competitors today, and be well positioned for reaping even bigger rewards of AI’s promised tomorrow.

[@anandkumarh7439]

excellent. Very good. would like to see such more case studies industry specific and over view of AI implementation.

[@harshithak7059]

Is there good open data available to work with or is everything hoarded

[@ruuubbbeeen]

thanks a lot for your videos Siraj. I've been learning a lot thanks to you.

[@sujitmohanty1]

Nope , not quite. U ve to break it down ....its tougher then it seems ..AI != AGI !=supply chain.

[@CuddleStories]

Saraj, Can you cover how we can do the Q/A ChatBot? I currently work in procurement and it would be very useful.

[@larryteslaspacexboringlawr739]

hi siraj, thank you for reading comments, do you know any ai systems that are really good at explaining themselves to ignorant managers with pretty pictures and easy to understand charts? i think many technicians, engineers, and programmers would love to have any ai products that help them avoid wasting their time educating non-technical supervisors. i see news head-lines like 'ai is a mysterious black-box' so any ai product that can visualize its decision process or feature bias or judgment system would be super sweet right? thank you for awesome videos

[@SirajRaval]

heres one for cnns github.com/merantix/picasso

[@pravinvedurla6787]

4:15 Homies for life mate. 😭😭

[@Naufal331]

Yo siraj Would love a video explaining the intuition, mathematics and application of the new and awesome ML algorithm XGBoost and how its different from the regular gradient boosting class in sklearn.ensemble.

[@KingPyRo1000]

Would love to try it, been using keras quite a lot but what scale of a dataset would be seen as relevant?

[@diegomdex]

@Siraj Raval, how we can be your fellows or apprentice in order to master Data Science (in my case for Supply Chain)?

[@AbhishekKumar-mq1tt]

Thank u for this awesome video , I need for more video on this topic

[@samratkokkula5558]

I cannot find the github link to the LSTM_forecast_year.py. Can someone please share the link with me

[@navneetmadhukumar3496]

The github link does not actually contain the keras script ???

[@anujbeatles]

Is there any way you could do something covering all of these in Salesforce?

[@mohamedaliaitnasser9455]

Good video , could you please do a video about predicting user behavior in an ecommerce website or any website it will be nice and and helpful thanks

[@florijnterstal6590]

I love what you are doing with these serie's i have always the problem what can i do with IA... so please do this it give great idea's ! Suggestion can you make one of the pulp and paper industry. To make paper there are a lot of variable's the big problem is getting the right color of the paper. Because it's a natural product you always need adjust the process bij hand. The time to get it on color could be redused by AI but not sure how... I think there are about maybe 5000+ variables..

[@yaghiyahbrenner8902]

IBM Watson sucks. But your video is very good and makes it look good.

[@SirajRaval]

its ok. thanks

[@forfreedomssake4315]

GENIUS thank you buddy ! :)

[@sagardilwali6866]

Siraj is da BOSS!! thanks man

[@statyoutube]

Hi Siraj, how can I get this data for my analysis.

[@danielniels22]

Hello, i'm also trying to work on this, but lack of data. Have you figured that out??

[@jessedampare1379]

AI for energy!!

[@rgrimoldi]

My favorite video so far!!!

[@ABOUZARWALA]

Really interesting stuff. Thanks a lot siraj. By the way I am for sirajcoin.

[@jacobvarghese7450]

IBM has blocked the public access on github? Any other way to get the source?

[@djsvideodiarys]

AI will be a curious psychopath by definition.

[@keshavlakhotia1432]

One more good concept

[@0dyss3us51]

I like the real Siraj :)

[@priyanshusangal9305]

How can we use it in Bench management?

[@sanyamjain4326]

Where can i find the code shown in last video

[@maikbanner7552]

It would have been good if you could have run the Full Cycle End to End from Data Transformation to Predictive Model Output.

[@akiforu]

Thank you Siraj..Amazing content

[@jusjosgra]

Did IBM pay you to make this video?

[@SirajRaval]

no

[@jusjosgra]

Thanks for your reply! I was just wondering since most of your other videos focus on open source systems like tensorflow.

[@navishkotwal]

Awesome stuff again . Thanks siraj ..

[@nickolay_startcev]

Великолепно! Всё очень доходчиво объяснил.

[@zachcase]

How about a video on AI for creating a budget to optimize a user to reach a financial goal

[@syjaf6615]

Love your videos. But dude!!! you've gotta do something about boom and echo in the audio... It gets sooooooooo annoying after 2 minutes!!!

[@prithvirajmaurya4538]

Siraj you're awesome ...great videos

[@rajankumar3733]

Does AI a threat for job in this sector?? Please reply

[@rufiromang]

Needle in the haystack? No way... it is much more complex than that. To find needle in haystack is easy. Just use magnet to do that... hehe...

[@revantkulkarni8734]

Hey great content! Can you please share the jupyter notebook? Thanks !

[@adirokr98]

Can you give the link of dataset?

[@DZT-ve2kx]

Your videos are awesome!!!

[@aa888zz]

I agree with most things except for the blockchain part. Blockchains DO NOT make good data bases. It would be way way better to just use digital signatures.

[@Roboticredemption83]

You are a frickin genius :)

[@iAM80tv]

Guy you’re really hilarious. Nice video

[@gadicherlapadmasri2276]

Can you please do a video on procurement specific?

[@ssrram525]

Please don't tell stories...come to the point...its totally waste and boring.

[@rebhuroy3713]

hey siraj you are one of the best driving instructor...will you please help me how can i start learning hacking...please be stuffs free...

[@larryteslaspacexboringlawr739]

thank you for supply chain video

[@Oswee]

It is right direction, but you are skipping huge part to achieve this.

[@poushalybd]

I would like to work for you, since I have a SCL background and currently learning ML for prediction, but it's mostly for learning from the best! Exploit me.

[@gabkov]

Nice vid

[@ganstabreakincity]

4:04 Literally!!

[@deepchandrakandpal3331]

very good video..

[@arthdh5222]

Video on 1D CNNs ?

[@narendranathreddy5046]

AI for contract analysis please? :)

[@zanderceo91]

siraj for the kids. secure the bag ha.

[@kmillanr]

Was hoping to see you clean that data

[@hamzahsajidkhan3602]

First comment :)

[@user-hm9ee5oh8l]

Hamzah sajid Khan congratulations bruh! 😄

[@tkdevlop]

old

[@spicytuna08]

how about some outputs?

[@qwertyplm13does51]

5:30 US Army in first gulf war

[@Naiiltwister]

Amazing!