How AI performance management startup Predera is helping enterprises deploy machine learning and automation at scale
San Jose-based Predera’s unified end-to-end automation engine provides intervention alerts, human-in-loop feedback, and autonomous workflow management capabilities to reduce the cost of maintenance of AI models.
Artificial intelligence, machine learning, and data scientists are together changing the world. Engineering teams work long hours experimenting with dynamic code and deploying them to achieve results. But it often turns out to be expensive and time consuming, especially with production models not available.
Now, San Jose-based AI performance management engine Predera is changing the way companies build, deploy, and monitor AI models at a lower cost. They enable data scientists to train AI for a business purpose (front-end application) rather than waste endless hours tagging and cleaning data at the database or infrastructure level.
Founded by Vamshi Ambati and Nazeer Hussain in 2017, Predera aims to bring “automation to machine learning so data scientists can do better data science”.
The two founders, friends from 2004, are technology experts who have worked in corporate for 15 years. Nazeer has worked at IBM and NTT Data, two large corporates pushing the boundaries in using data. Vamshi, a PhD in computer science from Carnegie Mellon, is a data scientist who has worked for the likes of PayPal.
In 2016, Nazeer and Vamshi decided to launch a company in Hyderabad and in San Jose to help other companies build their AI and machine learning strategies. As a services company, they realised that they could build a platform that manages AI for companies. And that’s how Predera was born in late 2017.
On Microland’s 30th anniversary, Kar has a visionary plan for technology to do more and intrude...
The startup is building the world’s first automation engine for managing cognitive apps. The product enables continuous integration, human-in-loop feedback, quality monitoring, and auto-updating capabilities required to reduce the cost of maintenance of AI models, which requires teams of data scientists as of now.
“We help companies manage their entire machine learning stack. They can run and train models on large scale data sets. Our platform is like an enterprise-grade data scientist working for a startup,” Vamshi says.
The founding duo invested $300,000 in the business, which includes personal money and the support of a few angels. Predera currently employs 15 people.
They are currently based in the FalconX Accelerator in Milpitas, which has been connecting them to corporate and VCs to scale the idea. They also have an office in Hyderabad.
How does Predera help?
Building AI from the ground up can be expensive if a company uses data scientists to clean and organise data. This is where Predera comes in, ensuring that data sets are available in the right format so that data scientists can crunch data instead of worrying about the organisation.
The platform ensures that teams never lose an AI/ML experiment, artifact, metrics, and lineage. It ensures that data is collected from all programming languages and that teams collaborate.
Predera offers a Smart Operations Engine that brings the complex art of taking experimental models to deployment at scale on a variety of infrastructure, all with one single click. It allows teams to monitor data science and machine learning models in production in a reliable, scalable, and explainable way, so data scientists spend less time debugging them.
With an integrated approach to managing model deployments, Predera acts as a dashboard and offers insights for model challenges, but can take actions on their behalf for fixing issues in production. The product works when integrated with libraries that host machine learning algorithms like TensorFlow, Pytorch, and Keras.
“Enterprises are just entering the era of AI and they need tools that can help them use data efficiently by deploying AI. Predera can show them how,” Vamshi says.
The market and the future
Global business value derived from artificial intelligence (AI) is projected to total $1.2 trillion in 2018, an increase of 70 percent from 2017, according to Gartner, Inc. AI-derived business value is forecast to reach $3.9 trillion in 2022.
The Gartner AI-derived business value forecast assesses the total business value of AI across all enterprise vertical sectors covered by Gartner. There are three different sources of AI business value: customer experience, new revenue, and cost reduction.
“AI promises to be the most disruptive class of technologies during the next 10 years due to advances in computational power, volume, velocity and variety of data, as well as advances in deep neural networks (DNNs),” says John-David Lovelock, Research Vice President at Gartner.
One of the biggest aggregate sources for AI-enhanced products and services acquired by enterprises between 2017 and 2022 will be niche solutions that address one need very well.
Predera competes with companies such as Dataiku, H20, and RapidMiner, which offer similar services.
The business model for Predera is based on the algorithms deployed from development and production to scale. This software-as-a-service company, which includes a global fintech credit card company and a couple of pharma companies among its clients, expects to scale its business in 2020.
Predera, which uses an integrated approach to provide a unified experience and put every data scientist first, is clear about its future plans: it wants to be one of those niche solutions that will go global.
(Edited by Teja Lele)