Robark co -founder Soham Mazdamdar, who left in 2023, has been called a new data startup Visit. The company offers AI data analtics that can provide business insights with structural, non -structured, and even “dirty” data, which means that the data is not cleared of the data or errors.
Where and how to work with the data, it is primarily Holly Grill for the Enterprise Business Intelligence Software and why Koto led a $ 23 million big seed round. Madona, GTM Capital, Anthology Fund, and others also participated.
Instead of asking the Data Analytics team to run reports, business manager can ask Wazoi’s questions and drill in details.
Mazdam presented an example of a chief revenue that would like to know, “How do I close my quarter?” The answer to the vasumi will submit a list of pending deals that should be focused on the team, as well as a list of questions waiting for each user, as well as delayed information.
“You can literally see the CRO through our platform, such as five major strokes, contrary to the process, including five people, and full -time,” Mazdar told Tech Crunch.
This is just an example that Wisoomai expects to respond.
Another initial customer is an oil and gas company that has thousands of workers in the field who use vsdomai to ask questions about production, taping data from everything stored to telemetry.
Taxkarnch event
Berkeley, ca
|
June 5 June
The book right now
Obviously, every business analytics device is already available-and a host of startups is also offering AI-powered natural language indications.
Wisdomy is standing for the breed of founders – they all first work with Mazdar in Robark. But the platform’s superpower is its accuracy, even against dirty data. It can find answers in the data stored in the files as well as the data, along with the data.
The important thing is that, Woodsumi will not supply deception.
Most businesses are looking for their AI app accuracy by focusing on the data used for collective generation training (RAG), such as model size, quick engineering, and, perhaps, for the model size, quick engineering, and, perhaps, real -time recovery techniques. Yet they are at risk of fabricated answers.
Wisdomi uses Genai to create a question – not in the creation of answers. “Finally, Geni can deceive,” says Mazdar. What we use to do is to write small programs … which can inquire from these different systems. ”
So if the model of Woodsuomi is hallow, it will write all a fake question that fails to recover the data. Self -data – Answer to the question – will not fabricate.
Visdomai has claimed to be the initial users of Konko Philps, Cisco, and Displash, and they have users working with large cloud data storage services such as Sno Flack, Google’s Big Curi, Amazon’s Red Shift, Data Berks, and Postgrams. Mazamdar says that by studying the language of inquiry and other sources, it can be trained on any data storage system.