As more online brands look for ways to move beyond third-party cookies as a way of gaining more direct insights about their users and customers, a startup that has developed a platform to help them has raised a big round of funding. Bluecore, a marketing technology firm that uses data gained from direct marketing like email, social media, site activity and combines that with machine learning to make better predictions about who might want to buy what among its customers, is today announcing that it has raised $50 million. The funding will be used to build the next generation of the Bluecore platform, expected later this year, which will tap into aggregated engagement data (but not actual browsing individuals) from "hundreds" of brands, which customers can combine with their own first-party data -- based on consent-based, first-party customer IDs -- to develop better targeting insights. "There are a lot of systems that focus on customer data and transactional data but no system that focuses on the product and product catalogue, which we think is the key asset," said Fayez Mohamood, the co-founder and CEO, in an interview. He says that the company manages over 200 million products and SKUs, second only to Amazon's and bigger than Walmart's, that companies can matches with consumer identities (from email and other direct channels).

Right now, there is no common ground with most of the AI technology consumers engage with. Instead of knowing who you are by simply the sound of your voice, most people are identified as a fresh user every time. Logging you in, authenticating you and knowing your history will be critical--and it all must be done seamlessly. Pins, passwords, security questions won't cut it,

A key factor in developing high performing machine learning models is the availability of sufficiently large datasets. This work is motivated by applications arising in Software as a Service (SaaS) companies where there exist numerous similar yet disjoint datasets from multiple client companies. To overcome the challenges of insufficient data without explicitly aggregating the clients' datasets due to privacy concerns, one solution is to collect more data for each individual client, another is to privately aggregate information from models trained on each client's data. In this work, two approaches for private model aggregation are proposed that enable the transfer of knowledge from existing models trained on other companies' datasets to a new company with limited labeled data while protecting each client company's underlying individual sensitive information. The two proposed approaches are based on state-of-the-art private learning algorithms: Differentially Private Permutation-based Stochastic Gradient Descent and Approximate Minima Perturbation. We empirically show that by leveraging differentially private techniques, we can enable private model aggregation and augment data utility while providing provable mathematical guarantees on privacy. The proposed methods thus provide significant business value for SaaS companies and their clients, specifically as a solution for the cold-start problem.

Way back in the last century, one of the most common put-downs of computers was the accusation that they only did what they were programmed to do. These days, it is increasingly common for marketing and many other kinds of systems to employ some variety of "machine learning," which moves away from the days when programmers dictated computers' every move. In this article, part of our MarTech Landscape Series, we look at this increasingly popular form of computing intelligence. "Historically," Bluecore CEO and co-founder Fayez Mohamood pointed out, "people wrote programs that were rule-based." His company is an email and marketing personalization platform.