Contact us

The future of process automation

02 Oct 2018 - Mickey Farrance

This podcast features Miguel Valdes Faura in discussion with Neil Ward-Dutton about the future of process automation

This podcast by MWD Advisors features Bonitasoft CEO Miguel Valdes Faura in discussion with Neil Ward-Dutton about the future of process automation.

Digital technologies are changing work and the workplace faster than ever before.

In this 35’14” episode, Neil and Miguel talk about BPM’s past and present, and where process automation is going next. They spend some time on support for different types of work, the impact of microservices, as well as (of course) the impact of Robotic Process Automation.


Some highlights:


Neil Ward-Dutton: What are you seeing in terms of changes in what people are doing?

Miguel Valdes Faura: I’m seeing, over the last two years, a major evolution in the nature of projects that is impacting the “what” and the “how” of BPM implementations.

Business automation and digital transformation initiatives are driving the change on the “what” a BPM project is, as customers are moving from more traditional operational projects - where the goal is to reduce cost, increase margins and improve efficiency of employees - to more strategic, innovative projects to launch new products and services that are oriented to customers, where the goal is to increase revenue and increase customer satisfaction.

“How” those projects are implemented is also changing. Digital transformation is all about incremental, iterative change, getting feedback from the end-user/customer fast, and measuring what is working and what’s not. Vendors are evolving to make sure that BPM technology allows innovation teams to do continuous delivery and continuous deployment of applications.

NWD: What kinds of Case Management use cases are you seeing when it comes to more exploratory styles of work?

MVF: Generally we are seeing three different use cases:

  1. Data-centric use cases or case management dealing with unstructured processes, that are usually related to support operations such as credit card fraud litigation service in a bank, or claims management in an insurance company. In the majority of these use cases, there is a need for more traditional BPM structured capabilities and processes. For example, reimbursement of the fraud to the customer is a well defined process.
  2. Exploratory styles of work use cases that end up being immature projects, because goals and use cases are not well defined. In these cases it's not about the technology.
  3. Use cases involving multiple companies or organizations in which the internal processes are structured but the collaborative processes between organizations are more unstructured and data-centric. For example in the US pharma industry, each pharmaceutical company can have a highly structured process to manage tests on potential side effects of a particular drug. But as soon as one of those companies detects a number of occurences of a side effect that goes beyond a certain limit, they must declare this to the FDA in the US. At that point a more unstructured, data-centric process starts, in which experts from all over the country will work together to understand the standard deviation.

NWD: Looking specifically at the trend towards use of microservices architecture, what do you see as the role and value of orchestration in that context? 

MVF: Everybody talks about the advantages of using micro services architectures and leveraging loosely coupled services accessible through APIs that can be created and managed independently. But of course there is a hidden complexity in creating those architectures: 

  • How to ensure reliability of each microservice
  • How to find the root cause of a problem I such a decentralised architecture
  • How to manage devops related complexity

I really think that orchestration engines, and even more, engines that support the BPMN2 standard, can help to manage asynchronous communication and things such errors, compensation and timeouts in mircroservices archietcture.

Some people will argue that adding an orchestration lawyer in a micro services architecture goes against the essence a descentralized architecture. I understand that, but we can also imagine a different architecture in which the orchestration engine is used inside each micro-service, combined with other technologies like a distributed streaming technology such as Apache Kafka for instance. Kafka will manage communication between services and the orchestration engine manages the long-running processes.

NWD: Also of course RPA is something that lots of organisations are investing in. We're seeing that as organisations get more established in their work with RPA, they're starting to look more at how automation can add value above the level of individual tasks. What are you seeing in this context?

MVF: The good thing about RPA technology is that it not only brings immediate savings to enterprises but it also frees up employees to work on more creative tasks.

Enterprises start to realise that they can use BPM and RPA combined to manage end-to-end process automation in their digital transformation initiatives. The BPM is responsible for process orchestration, customer interactions, exception management and complex decision-making by humans. The RPA is in charge of repetitive sequences of tasks that can be fully delegated to a virtual “digital workforce” of software robots. 

We really can see robots executing repetitive tasks as actors that are part of a global business process that is automated with BPM. In this use case BPM will be coordinating people, machines, things and robots.

For example, in projects related to customer experience improvements such as online customer account creation in which the process is managed by the BPM technology which also coordinates different interactions with robots (asynchronously) to get information about a prospect from different systems, trigger a robot to scan, review and  validate a document sent by the customer, and so on.

NWD: Can one orchestration engine/platform reasonably be used to address them all - or are there specialisations / optimisations that mean you need more than one specialised engine?

Some engines and platforms are more versatile than others but I really think that there is “no one solution fits all” when it comes to automation. Today we discussed only some of the complementarity between BPM, RPA and distributed steaming platforms.

That being said, if we look to what can be achieved with some of the orchestration engine/platform in the market these days, use cases include structured and unstructured scenarios, a mix of both, micro-services orchestration, digital customer experience creation where the platform also supports UI creation and the business data than will be connected to the business name a few.