The Human Resource industry is on the verge of a long-awaited transformation. The need for clear and well-defined processes make it an industry ripe for automation. Yet, considering the legal and human dimensions, it isn’t just about automating the customer-facing processes: it needs a complete overhaul. Language processing is rapidly evolving. We now have the tools required to design efficient HR bots to cover specific tasks.
I’ve recently met Arthur Boivin who is at the forefront of HR innovation to discuss these changes. The company he founded, Botmatic, is building the new tools of HR. It harnesses the power of AI to offer automated services that are tailored to each individual. We shared our vision and expertise and decided to co-produce a piece on the upcoming impact of chatbots in HR.
Impact of chatbots on human resources
Chatbots offer a unique advantage from an HR perspective: with a bot, what you ask is what you get. To put it another way: they remove the human variable in the enforcement of processes. There will be no personal interpretation or liberties taken with the process – if the process is sound so is the service.
One of the main concerns in HR departments is compliance. They require strict controls over the processes to make sure the company is compliant with employment and hiring laws. Data-protection is also a growing concern nowadays. The arrival of the new European data protection legal framework (GDPR, which will be enforced on May, 25th 2018) might be a turning point.
The main actors in the industry will have to sanitize their practices (at least in Europe…) if they want to avoid facing penalties in court. Indeed, the GDPR will completely change the way data is collected and handled:
Once the legislation comes into effect, controllers must ensure personal data is processed lawfully, transparently, and for a specific purpose. Once that purpose is fulfilled and the data is no longer required, it should be deleted.
Source: EU General Data Protection Regulation (GDPR) website
Chatbots will quickly prove themselves useful for GDPR compliance: they can help get consent, update information or offer accessible opt-out options for prospects. Essentially, they reduce the pain and the necessary steps required to update your customer, prospect or candidates files.
Introducing Botmatic: a new approach to bot building
Bots for Human Resources are not just another bot. The team behind Botmatic had to rethink the whole experience of creating a bot to match HR profile knowledge and habits. They launched an HR bot service built on the Botmatic conversational platform: HR Converse.
The UI is carefully thought through. Arthur Boivin, Botmatic’s CEO says their goal was to create a bot builder which would be “as simple as writing on Medium.”
In terms of UX, it translates into two main governing principles:
- Keyboard/Text only bot editor: the bot editor itself uses only text. Once a user gets used to the symbols and shortcuts, they can create bots really fast with nothing more than their keyboard.
- What you see is what you get: just like on Medium, the difference between the editor and the end-result are kept to a minimum. Users can try their bot live while seeing where they are in the flow in the editor panel.
A keyboard bot editor
The user experience on Botmatic is refreshing. The screen is divided into two panels:
- The “keyboard only” bot editor is on the left side of the screen
- The conversation tree is on the right side.
Both panels are in sync with each other: moving in the conversation through the text editor will switch the tree view to the right position.
Within the text panel, symbols and indentation are used to design the conversation flow.
- > indicate a user input | > Do you have any missions?
- : is a bot reply | : Yes of course ! When are you available to start?
- * creates a delay before your bot answer | *2s for a 2-second delay
Messages are nested together using indentation. To create a new user input with several variations, you can simply add all the variations (expressions) indented, below the input:
The creating and tagging of entities is also done from the text panel of the left. Any word can be linked to a property simply by clicking on it. On the screenshot below, “engineer” is being tagged as a “position”.
Property is the kind of expression the bot is waiting for. Entities on Botmatic are linked to SAP Conversational AI. No need to leave the Botmatic interface to go on SAP Conversational AI. They work together.
An impressive UX
The right side of the screen displays the conversation flow of the bot. It’s synchronized with the left side: clicking any box in the flow will bring you where the action is in the text, and it also works the other way around.
The test console is also synchronized with the tree:
Outside the bot building interface, Botmatic also provides features to segment your bot audience or create campaigns that are automatically sent when certain conditions are met. You can also handle the integrations, including ChatBase (bot analytics service) and SmartRecruiters (Applicant Tracking System) for now as well as the messaging channels: Messenger, Slack or Web Messenger. Skype will be available soon. The Botmatic team also deploy custom integration depending on the HR software used by its clients.
HR Converse bot
HR Converse, the first packaged bot made by Botmatic, tackles the recruitment and follow-up processes. Right now, applying to a company is often much more painful that it needs to be. The follow-up is often still lacking.
HR Converse is here to solve these issues with a chatbot. Integrated to all current channels, from emails to the applicant tracking system (such as SmartRecruiters), the bot is able to update the candidate depending on his application status. Since the update of the candidate’s profile is simplified, the information stays up-to-date and actionable.
The product proved itself particularly useful for recruitment agencies. Other specific industries, such as temp agencies enjoy the accuracy of the profile the bot builds. It learns the availabilities of each worker and keeps their information updated. This gives them a better visibility of their workers’ schedule.
Botmatic launched its early access a few weeks ago. Among the first use cases, some workflows are already finding their purpose: data collection and updating, availability management (temp agencies) and recruitment process follow-up.
From Dialogflow to SAP Conversational AI: building databases with conversations
To understand what were the needs and requirements of Botmatic, we went straight to the source: Samuel Roy, Botmatic’s CTO and co-founder.
“We used api.ai before it became Dialogflow. It was the first NLP API we used to design our prototype. Back in the day, we were already looking for a long-term partner since NER is key in our conversational platform.
On the support part, with the first integrations coming up, we needed to get in touch with the support team. Unhappy with the reactivity from Dialogflow at the time, we looked around, tried SAP Conversational AI and found our match. We had quick and generous answers from their support team to help us go deeper in our integrations. It was very much appreciated.
On the technical side, we also had better results with SAP Conversational AI than Dialogflow on the text recognition part even if now Dialogflow is bridging the gap. Indeed in early 2017, SAP Conversational AI had more entities available and more languages. More entities mean that we can deliver an accurate entity/property couple and in the end a better user experience.
On the integration part, it was not possible to create bots through the Dialogflow API. This means users would have needed to use two different interfaces to design their bot and plug it to the NLP. With SAP Conversational AI we built a full integration and reached a one-service solution.
In the end, we chose SAP Conversational AI as our key partner. We chose to integrate their NER solution because we needed to rely on effective support to deliver the best chatbots available 24/7. We were looking for a highly scalable service provider to grow the business.”
What’s next in HR automation ?
HR Converse is the first product based on Botmatic conversational platform and it will continue to expand. Indeed HR Converse is already used by recruiting and staffing agencies. It has been successful in screening candidates and keeping in touch with them long term.
The next step is HR automation with features like onboarding management. To deliver the best experience for new employees, companies can deploy 24/7 chatbots to provide safety rules, key information, registration checklist, key contacts and tips to help new recruits fit into their environment quickly.
As soon as onboarding is done the chatbot can take over and help employees manage their time off, timesheet, and overtime. They also take care of all the daily admin chores of corporate life. They’ll book your train ticket to Barcelona, and the room to go with it! Companies can get quick and easy feedback from their staff to monitor their happiness at work.
From recruiters to HR copywriters
Tomorrow, recruiters in charge of screening candidates and updating their profile on a regular basis will become copywriters. HR copywriters will be in charge of designing conversations to bring the best experiences possible to potential candidates.
Chatbots will free HR workers from the numerous tedious and repetitive tasks they are still doing today. Therefore, they will have the time to provide actual value to applicants. Instead of skimming through piles of CVs and taking care of the first steps of the recruiting process, they will focus on client and candidate relationships. They’ll be able to be more empathetic, advise applicants on their career choices and help them find the right training for the job they want.
HR people will spend more time on bringing the best experience to their talent pool rather than calling endless lists of candidates. In the end, empathy will be key and recruiters will be there to advise people on their career mobility. Chatbots will take care of the rest.