Rasa raises $13M led by Accel for its developer-friendly open source approach to chatbots

Conversational AI and the use of chatbots have been through multiple cycles of hype and disillusionment in the tech world. You know the story: first you get a launch from the likes of Apple, Facebook, Microsoft, Amazon, Google or any number of other companies, and then you get the many examples of how their services don’t work as intended at the slightest challenge. But time brings improvements and more focused expectations, and today a startup that has been harnessing all those learnings is announcing funding to take its own approach to conversational AI to the next level.

Rasa, which has built an open source platform for third parties to design and manage their own conversational (text or voice) AI chatbots, is today announcing that it has raised $13 million in a Series A round of funding led by Accel, with participation also from Basis Set Ventures, Greg Brockman (Co-founder & CTO OpenAI), Daniel Dines (Founder & CEO UiPath) and Mitchell Hashimoto (Co-founder & CTO Hashicorp). Rasa was founded in Berlin, but with this round, it be moving its headquarters to San Francisco with a plan to hire more people there in sales, marketing and business development; and to continue its tech development with its roadmap including plans to expand the platform to cover images, too.

The company was founded 2.5 years ago, by co-founder/CEO Alex Weidauer’s own admission “when chatbot hype was at its peak.” Rasa itself was not immune to it, too: “Everyone wanted to automate conversations, and so we set out to build something, too,” he said. “But we quickly realised it was extremely hard to do and that the developer tools were just not there yet.”

Rather than posing an insurmountable roadblock, the shortcomings of chatbots became the problem that Rasa set out to fix.

Alan Nichols, the co-founder who is now the CTO, is an AI PhD, but not in natural language as you might expect, but in machine learning. “What we do is more is address this as a mathematical, machine learning problem rather than one of language,” Weidauer said. Specifically, that means building a model that can be used by any company to tap its own resources to train their bots, in particular with unstructured information, which has been one of the trickier problems to solve in conversational AI.

At a time when many have raised concerns about who might “own” the progress of artificial intelligence, and specifically the data that goes into building these systems, Rasa’s approach is a refreshing one.

Typically, when an organization wants to build an AI chatbot either to interact with customers or to run something in the backend of their business, their developers most commonly opt for third-party cloud APIs that have restrictions on how they can be customized, or they build their own from scratch, but if the organization is not already a large tech company, it will be challenged to have the human or other resources to execute this.

Rasa underscores an emerging trend for a strong third contender. The company has built a stack of tools that it has open sourced, meaning that anyone (and thousands of developers do) use it for free, with a paid enterprise version including extra tools including customer support, testing and training tools, and production container deployment. (It’s priced depending on size of organization and usage.)

Importantly, whichever package is used, the tools run on a company’s own training data; and the company can ultimately host their bots wherever they choose, which have been some of the unique selling points for those using Rasa’s platform, when they are less interested in working with organizations that might also be competitors.

Adobe’s new AI assistant for searching on Adobe Stock, which has some 100 million images, was built on Rasa.

“We wanted to give our users an AI assistant that lets them search with natural language commands,” said Brett Butterfield, director of software development at Adobe, in a statement. “We looked at several online services, and, in the end, Rasa was the clear choice because we were able to host our own servers and protect our user’s data privacy. Being able to automate full conversations and the fact it is open source were key elements for us.” Other customers include Parallon and TalkSpace, Zurich and Allianz, Telekom, and UBS.

Open source has become big business in the last several years, and so a startup that’s built an AI platform that has a very direct application in the enterprise built on it presents an an obvious attraction for VCs.

“Automation is the next battleground for the enterprise, and while this is a very difficult space to win, especially for unstructured information like text and voice, we are confident Rasa has what it takes given their impressive adoption by developers,” said Andrei Brasoveanu, partner at Accel, in a statement. “Existing solutions don’t let in-house developer teams control their own automation destiny. Rasa is applying commercial open source software solutions for AI environments similarly to what open source leaders such as Cloudera, Mulesoft, and Hashicorp have done for others.”


By Ingrid Lunden

Salesforce and Google want to build a smarter customer service experience

Anyone who has dealt with bad customer service has felt frustration with the lack of basic understanding of who you are as a customer and what you need. Google and Salesforce feel your pain, and today the two companies expanded their partnership to try and create a smarter customer service experience.

The goal is to combine Salesforce’s customer knowledge with Google’s customer service-related AI products and build on the strengths of the combined solution to produce a better customer service experience, whether that’s with an agent or a chatbot..

Bill Patterson, executive vice president for Salesforce Service Cloud, gets that bad customer service is a source of vexation for many consumers, but his goal is to change that. Patterson points out that Google and Salesforce have been working together since 2017, but mostly on sales- and marketing-related projects. Today’s announcement marks the first time they are working on a customer service solution together.

For starters, the partnership is looking at the human customer service agent experience.”The combination of Google Contact Center AI, which highlights the language and the stream of intelligence that comes through that interaction, combined with the customer data and the business process information that that Salesforce has, really makes that an incredibly enriching experience for agents,” Patterson explained.

The Google software will understand voice and intent, and have access to a set of external information like weather or news events that might be having an impact on the customers, while Salesforce looks at the hard data it stores about the customer such as who they are, their buying history and previous interactions.

The companies believe that by bringing these two types of data together, they can surface relevant information in real time to help the agent give the best answer. It may be the best article or it could be just suggesting that a shipment might be late because of bad weather in the area.

Customer service agent screen showing information surfaced by intelligent layers in Google and Salesforce

The second part of the announcement involves improving the chatbot experience. We’ve all dealt with rigid chatbots, who can’t understand your request. Sure, it can sometimes channel your call to the right person, but if you have any question outside the most basic ones, it tends to get stuck, while you scream “Operator! I said OPERATOR!” (Or at least I do.)

Google and Salesforce are hoping to change that by bringing together Einstein, Salesforce’s artificial intelligence layer and Google Natural Language Understanding (NLU) in its Google Dialogflow product to better understand the request, monitor the sentiment and direct you to a human operator before you get frustrated.

Patterson’s department, which is on a $3.8 billion run rate, is poised to become the largest revenue producer in the Salesforce family by the end of the year. The company itself is on a run rate over $14 billion.

“So many organizations just struggle with primitives of great customer service and experience. We have a lot of passion for making everyday interaction better with agents,” he said. Maybe this partnership will bring some much needed improvement.


By Ron Miller

ServiceNow teams with Workplace by Facebook on service chatbot

One of the great things about enterprise chat applications beyond giving employees a common channel to communicate, is the ability to integrate with other enterprise applications. Today, Workplace, Facebook’s enterprise collaboration and communication application, and ServiceNow announced a new chatbot to make it easier for employees to navigate a company’s help desks inside Workplace Chat.

The beauty of the chatbot is that employees can get answers to common questions whenever they want, wherever they happen to be. The Workplace-ServiceNow integration happens in Workplace Chat and can can involve IT or HR help desk scenarios. A chatbot can help companies save time and money, and employees can get answers to common problems much faster.

Previously, getting these kind of answers would have required navigating multiple systems, making a phone call or submitting a ticket to the appropriate help desk. This approach provides a level of convenience and immediacy.

Companies can brainstorm common questions and answers and build them in the ServiceNow Virtual Agent Designer. It comes with some standard templates, and doesn’t require any kind of advanced scripting or programming skills. Instead, non-technical end users can adapt pre-populated templates to meet the needs, language and workflows of an individual organization.

Screenshot: ServiceNow

This is all part of a strategy by Facebook to integrate more enterprise applications into the tool. In May at the F8 conference, Facebook announced 52 such integrations from companies like Atlassian, SurveyMonkey, Hubspot and Marketo (the company Adobe bought in September for $4.75 billion).

This is part of a broader enterprise chat application trend around making these applications the center of every employee’s work life, while reducing task switching, the act of moving from application to application. This kind of integration is something that Slack has done very well and has up until now provided it with a differentiator, but the other enterprise players are catching on and today’s announcement with ServiceNow is part of that.


By Ron Miller

Oracle adds more AI features to its suite of sales tools

As the biggest sales and marketing technology firms mature, they are all turning to AI and machine learning to advance the field. This morning it was Oracle’s turn, announcing several AI-fueled features for its suite of sales tools.

Rob Tarkoff, who had previous stints at EMC, Adobe and Lithium, and is now EVP of Oracle CX Cloud says that the company has found ways to increase efficiency in the sales and marketing process by using artificial intelligence to speed up previously manual workflows, while taking advantage of all the data that is part of modern sales and marketing.

For starters, the company wants to help managers and salespeople understand the market better to identify the best prospects in the pipeline. To that end, Oracle is announcing integration with DataFox, the company it purchased last fall. The acquisition gave Oracle the ability to integrate highly detailed company profiles into their Customer Experience Cloud, including information such as SEC filings, job postings, news stories and other data about the company.

DataFox company profile. Screenshot: Oracle

“One of the things that DataFox helps you you do better is machine learning-driven sales planning, so you can take sales and account data and optimize territory assignments,” he explained.

The company also announced an AI sales planning tool. Tarkoff says that Oracle created this tool in conjunction with its ERP team. The goal is to use machine learning to help finance make more accurate performance predictions based on internal data.

“It’s really a competitor to companies like Anaplan, where we are now in the business of helping sales leaders optimize planning and forecasting, using predictive models to identify better future trends,” Tarkoff said.

Sales forecasting tool. Screenshot: Oracle

The final tool is really about increasing sales productivity by giving salespeople a virtual assistant. In this case, it’s a chatbot that can help handle tasks like scheduling meetings and offering task reminders to busy sales people, while allowing them to use their voices to enter information about calls and tasks. “We’ve invested a lot in chatbot technology, and a lot in algorithms to help our bots with specific dialogues that have sales- and marketing-industry specific schema and a lot of things that help optimize the automation in a rep’s experience working with sales planning tools,” Tarkoff said.

Brent Leary, principal at CRM Essentials, says that this kind of voice-driven assistant could make it easier to use CRM tools. “The Smarter Sales Assistant has the potential to not only improve the usability of the application, but by letting users interact with the system with their voice it should increase system usage,” he said.

All of these enhancements are designed to increase the level of automation and help sales teams run more efficiently with the ultimate goal of using data to more sales and making better use of sales personnel. They are hardly alone in this goal as competitors like Salesforce, Adobe and Microsoft are bringing a similar level of automation to their sales and marketing tools

The sales forecasting tool and the sales assistant are generally available starting today. The DataFox integration will GA in June.


By Ron Miller

Fresh out of Y Combinator, Leena AI scores $2M seed round

Leena AI, a recent Y Combinator graduate focusing on HR chatbots to help employees answer questions like how much vacation time they have left, announced a $2 million seed round today from a variety of investors.

Company co-founder and CEO Adit Jain says the seed money is about scaling the company and gaining customers. They hope to have 50 enterprise customers within the next 12-18 months. They currently have 16.

We wrote about the company in June when it was part of the Y Combinator Summer 2018 class. At the time Jain explained that they began in 2015 in India as a company called Chatteron. The original idea was to help others build chatbots, but like many startups, they realized there was a need not being addressed, in this case around HR, and they started Leena AI last year to focus specifically on that.

As they delved deeper into the HR problem, they found most employees had trouble getting answers to basic questions like how much vacation time they had or how to get a new baby on their health insurance. This forced a call to a help desk when the information was available online, but not always easy to find.

Jain pointed out that most HR policies are defined in policy documents, but employees don’t always know where they are. They felt a chatbot would be a good way to solve this problem and save a lot of time searching or calling for answers that should be easily found. What’s more, they learned that the vast majority of questions are fairly common and therefore easier for a system to learn.

Employees can access the Leena chatbot in Slack, Workplace by Facebook, Outlook, Skype for Business, Microsoft Teams and Cisco Spark. They also offer Web and mobile access to their service independent of these other tools.

Photo: Leena AI

What’s more, since most companies use a common set of backend HR systems like those from Oracle, SAP and NetSuite (also owned by Oracle), they have been able to build a set of standard integrators that are available out of the box with their solution.

The customer provides Leena with a handbook or a set of policy documents and they put their machine learning to work on that. Jain says, armed with this information, they can convert these documents into a structured set of questions and answers and feed that to the chatbot. They apply Natural Language Processing (NLP) to understand the question being asked and provide the correct answer.

They see room to move beyond HR and expand into other departments such as sales or customer service that could also take advantage of bots to answer a set of common questions. For now, as a recent YC graduate, they have their first bit of significant funding and they will concentrate on building HR chatbots and see where that takes them.


By Ron Miller

ServiceNow chatbot builder helps automate common service requests

When it comes to making requests inside a company for new equipment or to learn about HR policies, it can be a frustrating experience for both sides of the equation. HR and IT are probably tired of answering the same questions. Employees are tired of calling a help desk for routine inquiries and waiting for answers. ServiceNow’s new bot-building technology is designed to alleviate that problem by providing a way to create an automated bot-driven process for routine requests.

The company claims that you can build these bots to provide end-to-end service. Meaning if you tell the bot you need a new phone, it can pull your records, understand what you currently have and and order a new one all in the same interaction — and all within a common messaging interface such as Slack or Microsoft Teams.

It also works for customer service transactions to process routine customer inquiries without having to route them to a CSR to answer typical questions.

The new chatbot building tool called Virtual Agent, has been built into the ServiceNow Now platform and provides a way for developers to build conversational interfaces easily, says CJ Desai, chief product officer at ServiceNow. “[The Virtual Agent] enables our customers to develop a wide range of intelligent service conversations from a quick question to an entire business action through the messaging platform of their choice,” Desai said in a statement.

The announcement is part of a broader AI initiative on the part of ServiceNow, which purchased Parlo, a chatbot startup, just last week for an undisclosed amount of cash. The acquisition should help give ServiceNow more AI engineering talent and help them beef up their natural language processing (NLP) to further refine and improve their chatbot products moving forward, as the Parlo team and technology get incorporated into the ServiceNow platform.

The company claims that using these chatbots, customers can reduce call volume to help desks and customer service by 15-20 percent, using the standard argument that it should free humans to handle more difficult inquiries.

The company joins a slew of other platform players including Salesforce, IBM, Oracle, AWS, and others who are incorporating chatbot building technology into their platforms.


By Ron Miller

Google Cloud releases Dialogflow Enterprise Edition for building chat apps

Building conversational interfaces is a hot new area for developers. Chatbots can be a way to reduce friction in websites and apps and to give customers quick answers to commonly asked questions in a conversational framework. Today, Google announced it was making Dialogflow Enterprise Edition generally available. It had previously been in Beta.

This technology came to them via the API.AI acquisition in 2016. Google wisely decided to change the name of the tool along the way, giving it a moniker that more closely matched what it actually does. The company reports that hundreds of thousands are developers are using the tool already to build conversational interfaces.

This isn’t just an all-Google tool though. It works across voice interface platforms including Google Assistant, Amazon Alexa and Facebook Messenger, giving developers a tool to develop their chat apps once and use them across several devices without having to change the underlying code in a significant way.

What’s more, with today’s release the company is providing increased functionality and making it easier to transition to the enterprise edition at the same time.

“Starting today, you can combine batch operations that would have required multiple API calls into a single API call, reducing lines of code and shortening development time. Dialogflow API V2 is also now the default for all new agents, integrating with Google Cloud Speech-to-Text, enabling agent management via API, supporting gRPC, and providing an easy transition to Enterprise Edition with no code migration,” Dan Aharon Google’s product manager for Cloud AI wrote in a company blog post announcing the tool.

The company showed off a few new customers using Dialogflow to build chat interfaces for their customers including KLM Royal Dutch Airlines, Domino’s and Ticketmaster.

The new tool, which is available today, supports over 30 languages and as a generally available enterprise product comes with a support package and service level agreement (SLA).