RAIN Technology secures Series B funding poised to revolutionize sectors of the unemployed workforce with language solutions

By: Christos Makridis

RAIN Technology – a leader in speech and conversational AI with an agency services model and proprietary products – recently raised $11 million in Series B funding, bringing the total funding to $15 million. Investors in the round include Valor Capital, McLarty Diversified Holdings and Burch Creative Capital. The funding will allow RAIN to launch and scale its new product range alongside its existing agency offering.

RAIN’s first product hit the automotive sector last week – a custom assistant called “Ortho” that simplifies and improves the workflow of automotive repair technicians by allowing them to use their voice to access highly nuanced, jargon-heavy information in real time, without having to stop and use your hands to type. Building, launching and scaling this product is a big part of how recent funds are deployed.

Launching with voice in 2016, RAIN initially focused on consumer applications and mainstream voice assistants, partnering with companies and brands such as Nike, Amazon, Starbucks, Mastercard and Stanley Black & Decker to create “voice experiences” for smart speakers and mobile devices Apps and even the car. RAIN worked with nearly two dozen Fortune 500 companies and created more than 70 experiences with 15 million monthly interactions.

RAIN was an early adopter in language development towards custom assistants. There has been a real market shift where brands are looking to build their own first-party voice assistants in-house to have much more control over the customer experience and data. These are digital agents that operate via mobile devices, tablets, or custom hardware and provide a specific set of capabilities for an organization. Instead of building a voice app or voice experience on top of Alexa or Google Assistant, brands now want to cut out the middle man and let the consumer access their services directly, e.g. B. “Hey Nike” instead of “Hey Alexa” to control a Nike experience.

The unemployed workforce and compensatory behavior

Custom assistants aren’t just for consumers. RAIN’s business focus has shifted to developing people-centric solutions for an audience known as the “deskless worker” – the vast majority of workers who often have their hands occupied while doing their jobs and lack easy access to have a computer.

Over 2.7 billion workers worldwide – or 80% of the workforce – are considered “deskless”, concentrated in the agriculture, education, retail, hospitality, healthcare, manufacturing and transportation sectors. But they are underserved by technology, largely due to underfunding.

“Despite interest in increasing spending on deskless technologies, only 1% of software venture funding goes into technologies that serve 80% of the global workforce,” according to Emergence. The unemployed workforce is heterogeneous and varies in their work requirements and external work restrictions. Common characteristics of these jobs include:

  • The quality of work requires precision, efficiency, and sustained attention to detail without the “undo” button present in digital tasks.
  • The work is manual, complex and often performed in highly regulated environments.
  • A large amount of practical experience and mentoring is required to acquire expertise, often without formal mechanisms for cross-generational knowledge transfer.

“Workers in these industries cannot afford to make mistakes and give in to compensatory behaviors or ‘hacks’ over known challenges,” said Nithya Thadani, RAIN’s chief executive officer. “But these workarounds are often inefficient and present opportunities for innovation. We’ve seen these behaviors firsthand – construction workers sticking pins to their hard hats; Mechanics write vehicle data on the back of their hands; Peel technicians call up a patient’s blood pressure readings for someone else to write down.”

Speech technology can provide safer and more efficient solutions to these workarounds, but employers need tools that are both cost-effective and well-adapted to specific environments and workflows. Most of the existing tools were too generic to solve the end user’s unique needs and vulnerabilities, limiting their adoption.

AI-supported language technology as a supplement

Economists often look at automation and artificial intelligence as substitutes, at least for unskilled workers. For example, in a much-cited 2017 study, Carl Frey and Michael Osborne argued that 47% of all US employment is at risk of being displaced by automation. However, others have argued that while many occupations are exposed to automation and machine learning, few employees are fired outright.

Furthermore, the extension of AI, even if it displaces some tasks, can complement others and create entirely new types of work activities. “When AI focuses on augmenting humans rather than imitating them, humans retain the power to insist on a share of the value created. In addition, augmentation creates new capabilities and new products and services that ultimately generate far more value than just human-like AI,” said Erik Brynjolfsson, professor at Stanford University and director of the Digital Economy Lab.

These developments are also consistent with some recent empirical research on the impact of AI on well-being. For example, my recent research in Service Research Journal found that an increase in AI exposure in a city is associated with improvements in individual well-being, acting largely through an impact on local productivity. These results are robust to account for demographic differences across cities and to focus on changes in AI exposure and well-being over time. More research by me in Technological predictions and social change found that this result applies to even more general forms of technological change.

The specific data and computational requirements of machine learning models vary from industry to industry. For example, automotive repair professionals rely on multiple data sources to inform their work – databases that are structured differently on both the front-end user interface and the back-end. These sources can lead to different answers to the same questions, e.g. B. “Tell me the oil capacity of a vehicle” or “What is the torque for a wheel nut”. Choosing which database to consult and which specification to use can be far from easy and time-consuming for engineers. RAIN investigates how machine learning can help model technician decision-making processes so it can provide the “best” answer to repair questions by leveraging multiple data sources, saving technician time and improving the overall quality of their repair work.

“There are two overarching challenges to delivering on the promise of voice-enabled tools for the unscripted workforce,” Thadani continued. “First, the underlying data must be organized into meaning maps for a specific industry domain, known as taxonomies and ontologies; These allow natural language queries to be quickly and accurately parsed to retrieve relevant data and return it to the user. Second, speech technology must be built and tuned to work reliably in a real work environment, taking into account ambient noise and variations in users’ voices. Our goal is to develop applications that can hear and interpret natural language and industry-specific terms in a way that actually encourages usage and provides measurable value to professionals.”

looking ahead

While RAIN’s Agency Services Division continues to support enterprise-level brands in developing better solutions for employees and customers, the Product Division will continue to expand the use cases for Ortho, starting with capabilities for the automotive aftermarket user base and then exploring adjacent sectors such as trucking and other complex mechanical systems.

RAIN also works in other deskless vertical industries, including healthcare, with a strong need for purpose-built voice assistants. They have partnered with one of the largest healthcare providers in the US to deliver a voice-first prescription management application that has the potential to extend voice solutions beyond patients and into clinical settings. Hospital workers—surgeons, nurses, and OR technicians—experience similar pain points and environmental dynamics as other deskless workers—by using their hands in highly technical industries where precision and attention to detail are critical to getting their job done well.

source

Leave a Reply

Your email address will not be published. Required fields are marked *