MLOps Research
Press release

RealityEngines.AI rebrands as Abacus.AI, announces Series A funding and general availability of the world’s first autonomous AI service.

The company is announcing $13M in Series A funding and adding Mike Volpi from Index Ventures and Ram Shriram to its Board of Directors. In addition, the company will be open-sourcing a module that helps de-bias AI models.

July 14, 2020 - Abacus.AI, formerly RealityEngines.AI, an AI research and AI cloud services company, is announcing today the general availability of the world's first fully autonomous AI service. As part of the launch, the company is open-sourcing 3 novel techniques to de-bias algorithms and is introducing a new model showcase service. The new service will allow developers and data scientists to share and compare deep learning models and metrics, allowing businesses/enterprises to easily build cutting-edge real-time deep learning systems that can make a 5-15% impact to the bottom line. The company is also announcing $13M in Series A funding led by Index Ventures.

Bringing its total funding to $18.25M, this round was led by Mike Volpi from Index Ventures, with participation from Eric Schmidt, Ram Shriram, Decibel Ventures, and Jerry Yang. Following the investment, Mike Volpi and Ram Shriram will be joining the company's Board of Directors. Abacus.AI is going to use the funds to scale the service, gain early customers, and grow its research team.

Volpi is a veteran infrastructure and cloud investor and sits on the boards of Aurora, Cockroach Labs, Confluent,, Elastic, Scale, and others. Ram Shriram sits on the board of Google and participated in this round after co-leading Abacus.AI's seed round with Eric Schmidt, former CEO and Chairman of Google. Mariam Naficy, Erica Shultz, Neha Narkhede, Xuezhao Lan, and Jeannette Furstenberg, prominent women industry leaders, and Chet Kapoor, also joined in the round.

Abacus.AI helps organizations plug and play state-of-the-art deep learning systems into their existing customer experiences and business processes. The company's research team has invented a technique that finds the best neural architecture given a particular dataset and use-case. Once the neural architecture is identified, the system trains the neural net model and provides a simple prediction API that customers can use to embed predictions into their apps/websites. The system takes care of all the heavy lifting including setting up data pipelines, model re-training, and providing explanations for models.

During the beta phase, Abacus.AI worked with 1500 beta testers and tested hundreds of datasets to refine the accuracy of the deep-learning models created by their AI engine. The company is also launching a model showcase feature where users of Abacus.AI services can share their models with the rest of the AI and data science community. As part of the model showcase feature, Abacus.AI has published 20 models trained on public datasets. Accuracy metrics on these models are comparable to those from models that have been hand-crafted by data scientists.

Ensuring the AI models don't exhibit bias is as important as model accuracy. The company's research wing has invented three new techniques to de-bias AI models and will be open-sourcing its de-biasing library as part of this launch. The library can be used to measure and remove bias in models, regardless of how they were trained.

Several customers are already using Abacus in production and are already seeing improvements to the bottom line:

"Even a 5% accuracy improvement in demand forecasting predictions results in a significant impact to the bottom line of supply chain and manufacturing companies like ours," said Gus Shahin, CIO of Flex. "Abacus.AI's deep learning models have better accuracy than the other techniques that we have applied to-date and provide forecasts even if there is very little historical data."

"Abacus.AI's service empowers all our developers and data scientists to rapidly create powerful deep learning models at scale, in production. We can optimize all aspects of our user experience including personalizing emails, predictive churn, and providing contextual real-time recommendations. This translates a lift in both user-engagement and revenue," said Amit Shah, CMO, 1-800-Flowers.

"DailyLook, a premium styling, and fashion service uses Abacus.AI's services to create deep learning models to understand our user preferences and assist our stylists in picking the right outfits for our users. The models have helped increase engagement by 20% and reduced inventory headaches significantly," said Eric Marston, CTO, DailyLook.

Mike Volpi, Ram Shriram and Marty Chavez on why they choose to join the Abacus.AI team:

"We are excited to announce our Series A investment in Abacus.AI. As we've spent the last few months working with them, we are even more convinced that their vision of democratizing AI is spot on, and that they are going after a massive opportunity. Bindu Reddy, Arvind Sudararajan, Siddartha Naidu, along with their extraordinary team, are the right people to pull this off and the entire Index team is honored to be supporting their journey," said Mike Volpi, Partner at Index Ventures.

"Most organizations haven't been able to reap the benefits of AI/ML because it takes an army of specialized engineers and scientists to put deep learning systems into production. With innovative techniques like generative modeling for data augmentation and neural architecture search, the plug and play Abacus.AI service allows organizations to nimbly deploy autonomous deep learning models and instantly realize ROI," said Ram Shriram.

"I'm thrilled to join Abacus.AI as a key advisor as they prepare to launch the world's first fully autonomous AI service. I'm passionate about building software models of such high fidelity that they actually become the business or scientific process. Abacus.AI enables that kind of modeling, and the set of business applications is incredibly broad. Bindu Reddy is a brilliant innovator, and I look forward to being part of this team," said Marty Chavez, former CIO, CFO and global co-head of Securities at Goldman Sachs.

About Abacus.AI

Abacus.AI enables organizations of all sizes to easily plug and play state-of-the-art deep learning systems into their existing customer experiences and business processes without data hassles and complex engineering.

Abacus.AI was founded by Bindu Reddy, Arvind Sundararajan, and Siddartha Naidu. Bindu Reddy is the CEO of the company. Before co-founding Abacus.AI, she was the General Manager for AI Verticals at Amazon Web Services. Previously, Ms. Reddy was head of product for Google Apps before she started another company called Post Intelligence, which was acquired by Uber. Arvind Sundararajan, CTO of Abacus.AI, was a senior engineering leader in Uber's autonomous vehicle program (ATG) and at Google, where he was the engineering lead for the Gmail backend and AdSense's matching and serving system. Siddartha Naidu, Research Director of Abacus.AI, is the founder of BigQuery, the large scale serverless cloud data warehouse service offered by Google's cloud platform. Siddartha worked on a number of machine learning services and had senior engineering positions both at Google and Amazon. Abacus.AI is funded by Index Ventures, Eric Schmidt, Ram Shriram, Khosla Ventures, Paul Buchheit and others.

About Index Ventures

Index Ventures is a San Francisco and London-based international venture capital firm that helps the most ambitious entrepreneurs turn bold ideas into global businesses. Index-backed companies that are reshaping the world around us include Aurora,, Instabase, and Scale.

About Ram Shriram

Ram Shriram is an American investor. He is a founding board member and one of the first investors in Google. He serves on the boards of Google, Paperless Post, Yubico and Zazzle. He earlier served as an officer of working for Jeff Bezos. Shriram came to when they acquired Junglee, an online comparison shopping firm of which Shriram was president. Before Junglee and Amazon, Shriram was a member of the Netscape executive team, joining them in 1994 before they shipped products or posted revenue.