March 3, 2017

Running a business in the cloud - one year on

This time last year I wrote a blog post entitled "Running a business in the cloud" describing how we at Eagle use various cloud services in the day-to-day running of our business. At the time that post was written, we hadn't been using the various services for long, so I thought that now would be a good time to review how we've been using them, and mention a few other services we've started using in the meantime.

All of the services mentioned in last year's blog are still central to our operations. It's useful to look back and discuss what we did before we started using them.

Email / calendar

Our email and calendar were based on mostly self-administered IMAP and CalDAV servers, which led to a significant management overhead just to keep them running, and a selection of "interesting" compatibility issues with various clients. Email and calendar account management was separate, and we had to handle it all ourselves. Now, with Google Apps for Business, everything is integrated and we get much more power with virtually no management overhead.

Time tracking

Time tracking and billing was done with a home-grown system based on spreadsheets. They were tricky to fill in and keep up to date - we use a project-code based classification system, so every time a new project came along  the template spreadsheet had to be updated, and everyone had to make sure to use the same one. Reconciliation of the submitted time sheets was very time-consuming, and any budgetary analysis had to be done from scratch. Now, using Beebole, entries are submitted via a responsive, easy-to-use HTML5 interface, approval is trivial, and there are a wide range of built-in tools for tracking, analysis and reporting.

Holiday booking

Again, this used to be done with spreadsheets, which made requesting, approving and tracking holiday harder than it should be. Adding allowances when the new leave year started, carry-over and any special leave involved tedious manual entry and checking. Now that we use Appogee the process is greatly simplified, and copes automatically with differing leave allowances, pro-rata allowances for people who start mid-way during the leave year, end-of-year carry over and so on.


Our old documentation system was a self-hosted MediaWiki installation which was somewhat disorganised (admittedly our fault!), and so wasn't used as much as it could have been. We now use Confluence (as part of Atlassian's hosted OnDemand service) which has become central to how we work. Every project has its own Confluence space, and it's also the basis of our ISO27001 information security management system. There is also a very flexible permissions model, meaning that we can share certain project-specific pages with the customer for whom we are doing the project, which has really helped in terms of customer engagement and visibility.

One of the key things about all of the services above is that they use our Google Apps accounts, so there are no separate usernames and passwords to manage.

In the next part of this article, I'll describe some of the services which we have started to use since last year's blog post was written.

Topics: Big data technology, Bioinformatics