How our deep technical bench makes us better innovation partners.
Hi, I’m Dave, the CTO at Arctiq. One of my responsibilities is to make sure we’re transforming our delivery practice to be more effective with our large clients. This is a good glimpse at how we turned “David vs Goliath” to “David AND Goliath”.
We’re involved in a pretty darn neat HashiCorp Vault implementation at one of Canada’s top 10 largest organizations. The plan is for Vault to be the first tenant application sitting on an architecture that spans four data centers and Google Cloud. We own the architecture and deployment of a secrets and security framework that sits across those data centers as much as a firm of our size can own anything at a large national organization.
Things are going well (dare I say, swimmingly?) largely because we’ve been embedded into their organization. The request from the project’s Executive Sponsor was that, “we worked as one team.” Anyone in services knows many-a-client has made that request, and few have succeeded in blurring the lines between “we” and “they”.
We’ve figured out how to make it a reality, and the fact that we’re doing it as a firm of our stature makes me really proud.
To put it bluntly and plainly, the key was to map installation paths to decisions and communications plans. It’s not that hard of a problem to solve, but it relies on appreciating waterfall for what it truly is, and sneaking agile in to keep velocity high.
Arctiq is widely known as the go-to HashiCorp Vault partner when it comes to complex installations. We like the easy stuff too – but my team loves solving complex problems – and a five cluster security mesh delivering critical information across many data centers is exactly that.
At least it was a lot more complex the first few times we did it.
We’re comfortable enough with Vault to know the business outcomes and the proposed architecture are supported by the technology. Those of us that have worked in big companies – on big projects – know that it’s usually not the technology throwing wrenches into timelines and roadmaps.
So what causes the hiccups, false starts, hurry-up-and-waits that are so common in these big projects?
Surprises. That’s what holds projects up. Projects are run by people, and people hate surprises. What we’ve found is that because we know the tech so well, we can mitigate surprises. If we hear, “Can you approve this? The project can’t go forward until it gets approved,” then we know we’ve failed because someone needs to get involved that we didn’t identify.
(Note: I use “fail” here in terms of mitigating surprises – not the outcome of the project.)
So how did we mitigate surprises to keep things moving forward?
Well, being the best HashiCorp Vault partner with the best HashiCorp Vault consultants has its perks.
We were able to take the architecture and turn into a comprehensive implementation plan. We then took the implementation plan and looked at each line-item that needed a supporting decision. We relied heavily on our RAID log to track and monitor these decision points, being sure to identify which teams and individuals needed to be involved.
This is a framework we’ve applied in a few other large projects, and we think we have something very valuable. It sounds an awful lot like waterfall – and it is – because it relies on knowing what decisions need to be made based on what the desired outcome is.
Here’s a more concrete example:
Part of the installation involves choosing where logs should go.
We know that in simple installations with smaller teams that question can typically be answered at a director level, if not lower in the org. Not so much in large FIs.
Log targets for net new applications and platforms require serious planning. We knew we couldn’t rely on a five minute conversation to get an answer. We added the line to the decision log, and added some additional questions to ensure we had everything we needed.
Our client’s project coordinators were able to find out who we needed to speak to and organized a few meetings (totalling about six hours) to help us get to ground on what an inexperienced firm would estimate is a five minute chat.
Because we have that experience, we’ve been able to apply that kind of critical thinking to each decision point and include these kinds of things in our cost estimation. This makes fans of fixed price engagements happy.
We’re also working in parallel in a more agile manner to help define the artifacts that the next applications will need when they’re on-boarded onto the compute clusters. So it’s not a complete reversion to waterfall – it’s leveraging the good parts of waterfall in support of larger communications challenges.
We’re seeing similar success in the Apigee X and Hybrid projects we’re working on, so something about our framework is resonating with organizations that are leading innovation. As we put more and more successful projects behind us, we’ll be refining our framework and we’ll be sure to post about it.
What this really means to executive sponsors is predictability and confidence of task completion. When you marry our deep technical knowledge with a solid governance framework, we remove surprises from our projects.