Techseen: Is there any way to tell a good bot from a bad bot?
Aggarwal: Bots are basically the worker bees of software. A bad chatbot complicates the workflow instead of streamlining it. Good bots identify and index user data to determine the best ways to speed up processes that would otherwise be tedious or time-consuming. Bad bots consume bandwidth and slow servers down.
Much of chatbot security, and what makes a good bot, depends on human intervention at the end of the day. We have to develop bots that react accordingly when asked a question or confronted with unexpected content, so as to drive process improvement. Another issue with bots and machine learning is when a chatbot runs into a bad customer interaction and uses it to train itself.
This being said, filtering options for bots are still limited. However, platforms like Slack provide a way for team admin to add bots – which, again, assumes due diligence by the human admin.
Techseen: What are the different ways in which Ramco’s bots help in improving workplace productivity?
Aggarwal: Ramco’s JLTBot (“Just Like That” Bot) covers a whole range of employee self-service requirements through the bot. It’s like a personal office assistant. The point of bots is ultimately to streamline the digital workflow for everyday processes so that company time isn’t spent wading through menus. Contrast with how it usually takes an average of 60% repeat queries to HCM support to get a suitable answer, and up to 18 hours to get a response from HR.
Our bots are able to support functions that include but are definitely not limited to:
• Leave/time off (balance enquiry, new application, status, withdrawal)
• Expenses (status check, new application)
• Scheduling (timecard status, overtime, weekly schedule, off days)
• Payroll (payslip by month or year with a breakdown)
• Travel planning
Techseen: Unlike the web, which often provides at least a few signals that an interaction is secure, mobile apps could have hidden malware bots that could pose a significant threat to users. What kind of threats are there on the mobile ecosystem and what solutions does Ramco offer?
Aggarwal: For Ramco, the data goes through a fully secure internal bot platform – with various decision and dialogue engines – and then an API gateway that has to be authorized before it reaches the HCM database. All bot interactions on messaging platforms are more secure than web channels. Chat platforms can’t access devices to cause damage unless users download malicious attachments.
Techseen: Ramco has recently created about 34 bots related to human resources-related functions and for the aviation industry. What was the main purpose behind creating these bots?
Aggarwal: For aviation, Ramco’s bots were developed to cover functions such as part availability, AOG situations, customer relationships, stock status and more. A mechanic can gain visibility on all this data just by talking to a bot, which will respond by pulling up process videos or even suggesting and ordering components from a warehouse.
For HCM, as mentioned above, we cover employee expenses, travel, time off, schedule, pay, benefits and more. Typically, transactions which are recurrent and repetitive in nature are best addressed by bot automation.
Techseen: What kind of challenges did you face in integrating the bots with your existing software platform?
Technically speaking, the process itself is perfectly feasible. Ramco’s mobile-friendly platform, REST API, is highly capable and therefore does not pose much of a challenge to connect. The REST API framework is also capable of connecting to third-party data sources for an integrated experience.
But how do we ensure a smooth transition, active engagement and long-term user retention? By integrating with existing platforms and processes.
We had to think about the key pain points that our users faced and give them a reason to move systems. We changed the typical email and messaging user interface – which often requires shifting through many screens – to a one-time validation, self-service “zero-UI” system: something that users can manage with a simple exchange of texts. From a back-end perspective, it’s almost like a customer relationship management system.
Techseen: What measures does Ramco take to protect its botnets against denial-of-service (DoS) attacks?
Aggarwal: Ramco bots have strong internal and external security. They are locked to accept requests only from certain hosts or domains of messaging platform providers. It is therefore unlikely that other domains can reach the bot servers unless they break into the messaging platform provider’s networks, which are very secure.
Techseen: Do you think the increasing popularity of bots, especially chat bots, will have a similar impact on jobs?
Aggarwal: Jobs will likely not be lost. At most, the nature of jobs may change as elementary-level or front-office support jobs, like customer service, shift to more value-added ones with a focus on strategic work. This will make user experience better and mean better brand perception for businesses. Intuitive enterprise software can slash labor budgets and help skilled staff focus on higher-level work, which will be a boon in just about any business sector.
Techseen: Now that you have already successfully placed your bots in the aviation sector, do you plan to integrate your bots in other industrial sectors as well?
Aggarwal: We are currently experimenting with bots in logistics- and manufacturing related sectors. Based on business needs, we may expand this further.