Karthik Sajikumar's Profile

Absolutely Louis! I do think that macro economic conditions + a new nascent market has been the main reason for C3 AI’s stock to perform poorly. To yours and Gigi’s point – The biggest opportunity I see is being this ‘one stop shop’ for all your AI model needs and really providing a complete turnkey enterprise solution. The main competitive advantage that C3ai have right now is not only providing an end-to-end platform of solutions : supply chain planning and execution capabilities, with extension to demand planning, production scheduling, sourcing management, and supply network node risk, process optimization for improved manufacturing but the ability to deploy this at scale.

Granted these are probably not a competitive moat compared to a slew of competitors who are specializing and entering this space. Since it’s a very nascent and growing space, I do think that execution will define and determine the winners and losers in the market and then the management can really see the effect of some of these applications, what companies really want and then honing on those applications.

Coming from manufacturing and operations and as a person who have seen multiple downtimes due to equipment failures on assembly line, I cannot emphasize more on how important of a problem this is. The question still remains on the data you get from the equipment. Most of the large corporations have legacy systems in place that might not give you the right data that’s needed or sometimes can provide wrong information [ Since a lot of it depends on the sensors that monitor the equipment and assembly], the other issues is the hardware calibration that’s needed on a routine basis. Does Falkonry provide information on those sensor calibration and can take decisions based on the variances measured because otherwise a false positive is equally bad when there’s no issues on hand

The other question that I am trying to wrap my head around is how much of this data can be transferred and applied universally in building algorithms ? A robot which lifts a box of medical supplies from point A to B will have a different life cycle for maintenance compared to a similar robot which picks a much heavier object. How does Falkonry think about this dilemma? Do they create unique models for each customer or do they look at the models they previously have and modify them for this scenario?

Thank you Isabella for this article! This is a space I am so curious about. I remember seeing the Michael Jackson Hologram and really being fascinated by technology. I think by piecing together all these articles, the metadata about the customers and their listening patterns and countless hours of songs these legendary artists have created, you can potentially use AI to come up with a new album from some of the greatest bands/musicians ever lived ( including Beatles, Pink Floyd etc.) but then the question still remains – will it completely replace artists and their ‘secret sauce’ ?

I think another application that Spotify can get through this data is to look at who is going to be the ‘next big artist’. They have hours and hours of data of musicians and listening behavior ( when you forward, when you rewind, how many times did you play, what specific words/instruments triggered that?) – this can be used to scrape the internet and get new artists hired to Spotify platform – in-fact in a way becoming a studio

I really liked this article. Thanks a lot Manuel! I do have some questions regarding AI in factory operations. One of the most difficult things to in the defense and aerospace world is the access to data. I’m sure LM has customers outside the United States but are allies of the US. How does LM manage data security while also getting data from its end products ( Black Hawk, Drones etc.) to constantly improve the algorithms? and how does that affect the infrastructure that LM needs to invest in

I tried doing this and it’s unbelievable if you type “who’s going to win the election in 2024?” ( Hint : it shows the same person). In a way this if done correctly is a great tool for political campaigns. They’ll be able to understand the imagery out in the web pages and who’s leading on that end and what audience are seeing.. This can positively ( or negatively) direct their digital strategies maybe?

This is quite interesting! I wonder if things would be any different if you tried ‘holiday’ or ‘time off’ and it’ll all show people relaxing on a beach! I feel like when we think about these in our head – it’s almost a beach vacation that we all dream for? or at least the majority of us? Looks like AI is then doing a good job haha

Haha! Got to love this.. I, for some reason feels that AI is trying to say a story here if you go tile by tile.. starting from you looking or getting an idea to the last frame where your colleagues are probably looking upto you as a person who made it or as Isabella pointed out, it could be my interpretation

Thanks a lot Laura for this post! As someone who loves vine, I can appreciate a great app that can help curate my wine list whenever I need this. That organic growth numbers you posted are any platform’s dreams! Wineries will pay so much for this rich customer data so that really curate offerings to the customers that are buying their wines. The data they have at this point through retail channels and other outlets show only a fragment of who or what the customer is and so is not that useful compared to what this app can do for them

I love the idea of expanding to other alcohols and potentially can add videos or educational materials on the type of alcohol, the origin as well as some of the cocktails they can make out of it. What they could also expand into is to really use data from the reviews (250M reviews!) and give the users sophisticated recommendations on wine pairings as well as bringing in exclusive wine enthusiasts and sommeliers to provide a ‘featured list’ every so often to increase the frequency of customers visiting the app thereby leading to more revenue. Another source of increasing the stickiness of the customer, as well as growth, is by having virtual/ in person wine tasting events in partnership with a wine club

Couple of interesting questions that came to my mind while reading this are :

1) What kind of algorithm do they have to screen millions of bottles of wine – are they doing by themselves, who uploads these photos, when do they refresh this? what’s the accuracy rate/customer delight/other KPI’s on recommendations?
2) Since it’s a free app for users as well as sellers, how’re they screening for fake reviews and ensuring quality? If some sort of quality is not enforced, this could lead to a lot of customers and sellers leaving the platform

This is really interesting! I’ve never heard of TBL but really looks like a promising! I checked out their website and noticed that they don’t have that many reviews/tours ( Paris & London) which tells me that they’re either catering to the wrong customer segment or that their marketing campaigns need to be really targeted.

I think the real problem to scale and sustain this product on top of what you have mentioned is whether I have any incentive to go to this website : When AirBnB, Orbitz and even Marriott Hotels offer ‘experiences’, they’re really tapping into extensions of the offerings to get an increased share of my wallet while acting as a ‘one stop shop’ for all my travel needs.

I do believe that this model has a lot of merit but I’m really curious to understand who are they catering to vs who purchases their products and see if there’s a disconnect which can help them tweak their product better to the customers who wants these private, very local tours. That can then differentiate them by catering to a specific segment – maybe luxury segment where you want very private tours or groups who are on a budget etc.

Thanks a lot Natasha for this insightful article. Coming from a supply chain and operations background, I first hand understand this problem really well and the value proposition Warehowz has for a number of D2C SMB’s. However the challenge with this model I feel are two folds :

1) End to end integrations : The biggest logistical nightmare for any scaling up company ( solar panel mfg company in this scenario) is to understand the right levels of inventory, tracking of orders and have that integrated with their own sales, operations and production planning. The model here has you outsource a part of your supply chain – in this case, the warehouse/storage aspect of it – can lead to massive headaches later on. a) this model needs to assure that all these warehouses on the platform have certain criteria – accuracy in inventory management, right shipping solutions, packaging ( if needed) which can lead to very complicated operational challenges. b) Once you introduce a new product or a new ‘refresh’ on your existing product, you have to purge your old inventory systematically. If not, there are chances of having wrong products shipped to the end customer affecting the brand loyalty and value for the manufacturers and potential safety issues for customers

2) Software component and touch points: How do we make sure that the software the warehouse uses ( scan a product, keep it on a shelf, move it to dock area for transport and finally transportation) has a lot of touch points that needs to be properly entered into a database and work seamlessly with any operational tools the original manufacturers use. When you have multiple people renting your warehouse at the same time, it entails that the software Warehowz uses should be compatible with the manufacturer’s. This posts an interesting challenge that really needs some thought

I love this post! Thanks a lot Nthato for posting about F1!

Coming from a mechanical engineering and automotive background, I’ve seen first hand the importance of CFD when it comes to drag, performance and decision making on even coatings that goes into many of the components. The importance of those measurements grow 10x or even 100x when it comes to competition such as the F1 where split second decisions taken will change the outcomes drastically. It’s so fascinating to see how many reengineering iterations happens in an F1 season these days thanks to the data that’s created by these hundreds of sensors

I remember reading about Mercedes Benz team and the terabytes of data that they generate on a daily basis and how they were able to reduce the computational time required to make better decisions. At the same time, it reminds me of the TSG Hoffenheim case on what do you do when data becomes not a competitive advantage and other players are slowing closing the gap?

Thanks a lot for this blog post Jiwon!

I’m a Marriott loyalist and so loved reading about this article. Marriott, in my opinion, is a success story against disruption – in this case by AirBnb. One of the other things that I love about how Marriott how the Bonvoy app is equipped with experiences that customers can chose and use their points against thereby learning a lot more about the customer, their likes and dislikes and cater solutions to members in their own city even if they are not staying with a Marriott property – in a way of ‘owning the customer’ beyond the days where they are at the properties

Thanks for the post Nitya! I’ve always wondered on how we can put an effective strategy against Amazon. Be it in any verticals, Amazon is a threat. One thing I always wondered about in this space is the importance of ‘convenience’. With Amazon offering same day or 2 day deliveries or even books through Kindle, what’s the incentive for me to order a book through PRH? Maybe the answer is that I can utilize PRH for their recommendations and can use Amazon to order the book vs going through PRH which might take additional days for me to get the book? While PRH can utilize the customer data and run models on to understand customer behaviors, create personas and tailor recommendations and future titles to cater to that audience?