Daisho - Become a data science superhero, no code, no math

Hey Product Hunt! I am Swarna, the maker of Daisho, a no-code auto-ML platform for business users. Before starting to work on Daisho, I was your typical business guy working in sales, strategy, and operations for large as well as not-so-large companies. I was a numbers guy and Excel was my go-to tool. This worked well for basic analysis - I got all the low-hanging fruits. But after that, it was a slog. I needed data scientists and ML experts to get stuff done. This was expensive and time-consuming. There wasn’t a SINGLE tool in the market for people like me who didn’t know how to code! That’s when I reached out to Dipayan who was finishing up his Ph.D. in Statistics. We realized that there’s a huge gap between what users like me actually want and what’s out there. So we built Daisho for folks like me to solve business problems. You upload your data, answer a few basic questions, and that's it! Daisho does the rest! Over the last couple of years, we have had customers using Daisho to do tons of cool stuff and every time they learned something new! A business analyst for a healthcare service provider used Daisho to understand what was driving errors in their claims process. She had data over a few months with errors identified after an audit. All she did was upload his excel sheet, point to the column where the audit info was stored, and that's it! Daisho built a model in minutes and identified that errors were driven by associates not trained in a state being assigned claims in that state. This resulted in a rejig of their rules of assigning claims to associates. A product manager at an online jobs classifieds portal improved their conversion using Daisho. He downloaded their CRM data from SalesForce, uploaded it to Daisho, and pointed Daisho to the metric which tracks lead conversion. Daisho’s model picked up an anomaly in the data where deals expected to be mapped to a certain demographic were routed to a different one - just because the default settings in their site were using prefilled values when they should not have! Folks have used Daisho for predicting telecom network performance, impact of COVID on healthcare collections, yield management in manufacturing, and many more places! But, how does Daisho work across domains and contexts? How does Daisho know what’s important for you? It’s because Daisho is built on top of Signal Factory - Daisho automatically builds and tests 100s (and 1000s) of hypotheses for all domains - just based on the data - and filters the best of them. We got a couple of cool comics at https://www.daisho.ml/technology..., and https://www.daisho.ml/technology... that explain how! Today, we launch Daisho No-Code publicly with Driver Analysis and Time-Series modules - these are by far the most commonly used modules in our experience. We have a lot more lined up too! Have a look and try it out! I am here to listen to your feedback and help you all along the way!

Looks very promising) We are just about to process over 100 millions of pairs of investors and founders at our platform, covering over 70 parameters for each 🤯 and afraid that it can cost us a fortune😭 gonna show Daisho to my team (:

I think every data science team should try daisho - I have a feeling that >80% of everyday usecases can be easily solved here!

Thanks for sharing your story, Swarna! I love how it inspired you to created Daisho and so far I'm loving the logo, name, website, excited to give the free trial a go sometime! Great job in launching this!

We have launched with Business Driver Analysis & Business Time Series. We would love to see you try your data on Daisho! There's a ton of use-cases that we are planning to add to Daisho. You can log in and let us know what you think is important for you!

No-code driver analysis for businesses is really tough - you have to at least provide fully automatic data cleanup recommendations, figure out the hidden hypotheses in the data (coz, everybody knows that complex drivers are what drives a business), and finally showing very clear insights based on the predictive models for business users to make sense of it all! If figuring our drivers is your end-goal, thats a ton of ML automation when you start with just a csv file or an excel table! And Daisho does it all! While Daisho has successfully solved problems in CPG, healthcare, telecom, internet businesses, etc, I'd like to know where you think Daisho will be able to make the most impact.