bpmNEXT 2014 Wednesday Morning: Cloud, Synthetic APIs and Models

I’m not going to say anything about last night, but it’s a bit of a subdued crowed here this morning at bpmNEXT. Smile

We started the day with Tom Baeyens of Effektif talking about cloud workflow simplified. I reviewed Effektif in January at the time of launch and liked the simple and accessible capabilities that it offers; Tom’s premise is that BPM is just as useful as email, and it needs to be just as simple to use as email so that we are not reliant on a handful of power users inside an organization to make them work. To do this, we need to strip out features rather than add features, and reduce the barriers to trying it out by offering it in the cloud. Inspired by Trello (simple task management) and IFTTT (simple cloud integration, which basically boils down every process to a trigger and an action), Effektif brings personal DIY workflow to the enterprise that also provides a bridge to enterprise process management through layers of functionality. Individual users can get started building their own simple workflows to automate their day-to-day tasks, then more technical resources can add functionality to turn these into fully-integrated business processes. Tom gave a demo of Effektif, starting with creating a checklist of items to be completed, with the ability to add comments, include participants and add attachments to the case. There have been a few changes since my review: you can use Markdown to format comments (I think that understanding of Markdown is very uncommon in business and may not be well-adopted as, for example, a TinyMCE formatted text field); cases can now to started by a form as well as manually or via email; and Google Drive support is emerging to support usage patterns such as storing an email attachment when the email is used to instantiate the case. He also talked about some roadmap items, such as migrating case instances from one version of a process definition to another.

Next up was Stefan Andreasen of Kapow (now part of Kofax) on automation of manual processes with synthetic APIs – I’m happy for the opportunity to see this because I missed seeing anything about Kapow during my too-short trip to the Kofax Transform conference a couple of weeks ago. He walked through a scenario of a Ferrari sales dealership who looks up SEC filings to see who sold their stock options lately (hence has some ready cash to spend on a Ferrari), and narrow that down with Bloomberg data on age, salary and region to find some pre-qualified sales leads, then load them into Salesforce. Manually, this would be an overwhelming task, but Kapow can create synthetic APIs on top of each of these sites/apps to allow for data extraction and manipulation, then run those on a pre-set schedule. He started with a “Kapplet” (applications for business analysts) that extracts the SEC filing data, allows easy manual filtering by criteria such as filing amount and age, then select records for committal to Salesforce. The idea is that there are data sources out there that people don’t think of as data sources, and many web applications that don’t easily integrated with each other, so people end up manually copying and pasting (or re-keying) information from one screen to another; Kapow provides the modern-day equivalent to screen-scraping that taps into the presentation logic and data (not the physical layout or style, hence less likely to break when the website changes) of any web app to add an API using a graphical visual flow/rules editor. Building by example, elements on a web page are visually tagged as being list items (requiring a loop), data elements to extract, and much more. It can automate a number of other things as well: Stefan showed how a local directory of cryptically-named files can be renamed to the actual titles based on table of contents HTML document; this is very common for conference proceedings, and I have hundreds of file sets like this that I would love to rename. The synthetic APIs are exposed as REST services, and can be bundled into Kapplets so that the functionality is exposed through an application that is useable by non-technical users. Just as Tom Baeyens talked about lowering the barriers for BPM inside enterprises in the previous demo, Kapow is lowering the bar for application integration to service the unmet needs.

It would be great if Tom and Stefan put their heads together and lunch and whipped up an Effektif-Kapow demo, it seems like a natural fit.

Next was Scott Menter of BP Logix on a successor to flowcharts, namely their Process Director GANTT chart-style process interface – he said that he felt like he was talking about German Shepherds to a conference of cat-lovers – as a different way to represent processes that is less complex to build and modify than a flow diagram, and also provides better information on the temporal aspects and dependencies such as when a process will complete and the impacts of delays. Rather than a “successor” model such as a flow chart, that models what comes after what, a GANTT chart is a “predecessor” model, that models the preconditions for each task. A subtle but important difference when the temporal dependencies are critical. Although you could map between the two model types on some level, BP Logix has a completely different model designer and execution engine, optimized for a process timeline. One cool thing about it is that it incorporates past experience: the time required to do a task in the past is overlaid on the process timeline, and predictions made for how well this process is doing based on current instance performance and past performance, including tasks that are far in the future. In other words, predictive analytics are baked right into each process since it is a temporal model, not an add-on such as you would have in a process based on a flow model.

For the last demo of this session, Jean-Loup Comeliau of W4 on their BPMN+ product, which provides model-driven development using BPMN 2, UML 2, CMIS and other standards to generate web-based process applications without generating code: the engine interprets and executes the models directly. The BPMN modeling is pretty standard compared to other process modeling tools, but they also allow UML modeling of the data objects within the process model; I see this in more complete stack tools such as TIBCO’s, but this is less common from the smaller BPM vendors. Resources can be assigned to user tasks using various rules, and user interface forms are generated based on the activities and data models, and can be modified if required. The entire application is deployed as a web application. The data-centricity is key, since if the models change, the interface and application will automatically update to match. There is definitely a strong message here on the role of standards, and how we need more than just BPMN if we’re going to have fully model-driven application development.

We’re taking a break, and will be back for the Model Interchange Working Group demonstration with participants from around the world.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.