Enterprises benefit from real-time decision-making in four weeks versus four months.
Darshan and his leadership team all have big data backgrounds and are driven by the desire to help companies realize benefits from AI and ML more rapidly. Their mission is to accelerate organizational learning by an order of magnitude. They do this the same way Google did, by focusing on the feedback loop.
Isima created the biOS business intelligence operating system to serve API developers, BI consumers, and data scientists. The operating system is hyper-converged, intent-driven, and cloud-agnostic.
Enterprises have been stitching together SaaS and cloud solutions to aggregate disparate data to drive and create ML models. There is vendor lock-in with solutions providers which is great for providers and consultants but less so for the enterprise and customers seeking timely insights from their data.
The promise of hybrid multi-cloud and edge computing as a moat against Amazon and Google is very challenging and time-consuming even if you have the data scientists required to execute your vision.
Isima consolidates data so enterprises can benefit from the integration of data with a single source of truth. This provides time-to-value in days by enabling data scientists to build and tweak their models with real-time data ops, deployment freedom on IT-friendly architecture.
They work to solve the hard problems first with an order of magnitude of improved productivity. To date, use cases include: 1) real-time order fulfillment for a large e-commerce company; 2) real-time churn prevention for a major telco provider; 3) real-time trade reconciliation for an international trading desk; and, 4) real-time fraud detection for a major fintech lender.
Isima's average time-to-value has been 11.5 days, with one developer versus 60 days with SaaS, 90 days with a cloud PaaS, and 120 days with open source and anywhere from 7 to 30+ developers. This is the order of magnitude improved productivity to which they refer.
Darshan asks clients to give him their toughest data problem and their worst engineer and he’ll deliver a solution in four weeks as proof of how enterprises can use their data to make a quantum leap in production more quickly and cheaply than before.