Peak.AI, a startup developing AI solutions for enterprise customers, today announced that it has raised $12 million in extended series A funding. The fresh capital will fuel Peak’s growth, commercial expansion, and R&D, according to CEO Richard Potter, and will come as up to 25% of companies report experiencing a 50% failure rate in deploying AI models.

Despite the promise of AI, the corporate sector’s adoption curve hasn’t been as steep as some had predicted. A survey of publicly traded U.S. retailers’ earnings calls found that only nine of about 50 companies had started to discuss an AI strategy, and a separate study — from Genesys — found that 68% of workers aren’t yet using tools that leverage AI.

Peak aims to simplify implementation with a subscription-based software-as-a-service offering that spans infrastructure, data processing, workflow, and applications. Its customers — brands like Pepsi and Marshalls — supply their data, which Peak’s platform ingests through built-in connectors to accomplish things like optimizing supply and demand and supporting fulfillment processes, courtesy of a library of configurable AI engines.

Once AI engines go live, their predictive and prescriptive outputs can be exposed through APIs or explored, visualized, and exported with Peak’s Data Studio. The platform can handle data sets of virtually any size running on Amazon Web Services, and it serves models in an always-on fashion so that they self-improve over time. It also screens all ingested data through an algorithm to identify and anonymize any personally identifiable information.

VB Transform 2020 Online - July 15-17. Join leading AI executives: Register for the free livestream.

Peak’s team optionally works with customers to define objectives, quantify opportunities using a sample of data, and scope out a business case for sign-off and launch. It’ll take care of kick-off and onboarding, as well as operationalizing, and it’ll configure the solutions to individual user needs.

There’s no shortage of fully managed AI solutions with substantial venture backing. H2O recently raised $72.5 million to further develop its platform that runs on bare metal or atop existing clusters and supports a range of statistical models and algorithms. And Cnvrg.io — which recently launched a free community tier — has raised $8 million to date for its end-to-end AI model tracking and monitoring suite.

But Peak claims its platform is more performant than rival offerings. It says it has helped customers achieve a 28% uplift in marketing revenues, a 4 times increase in return on capital employed, and a 147-ton reduction in CO2 emissions through optimized logistics and resource planning.

MMC Ventures and Praetura Ventures led the series A round, which brings Manchester-based Peak’s total funding to $18 million. The company was founded in December 2014 by CEO Richard Potter, David Leitch, and Atul Sharma and has additional offices in Jaipur and Edinburgh.

Sign up for Funding Weekly to start your week with VB's top funding stories.