I help you choose where your product should think: in the cloud or at the edge.

The problem

Most IoT devices still just send data to the cloud and wait. The next generation decides on the spot.

The method

I guide IoT teams along the Edge AI Path pick the decisions worth moving onto the device, split the work between device and cloud, make the hardware you already have smarter, build in security from day one, in line with the EU Cyber Resilience Act, and roll it out across your fleet with outcomes you can measure. Every decision starts with the BLERP check (Bandwidth, Latency, Economics, Reliability, Privacy): if any of them matter, that decision belongs on the device.

The BLERP framework comes from Daniel Situnayake and Jenny Plunkett's book AI at the Edge.

Proof - guiding Azure IoT customers onto the Edge AI path

It’s the path I helped Fortune 500 teams navigate at Microsoft, from cloud to device.

• Customers could see at a glance whether their scenario was cloud-connected or edge-connected, and which product fit it.
• Customers left the Well-Architected Framework assessment with clear recommendations to make their IoT solutions more reliable, secure and cost-efficient, at the edge and in the cloud.
• More teams moved to Edge AI solutions.

The situation

“Should this run in the cloud or on the device?” Teams building on Azure IoT asked it constantly. The edge-connected pattern was their way into Edge AI, but many weren’t sure when it fit, so teams with industrial, low-latency or security-sensitive devices often defaulted to the cloud. Their questions didn’t always get a consistent answer.

What I did

As an Azure IoT subject matter expert at Microsoft, I worked on that question from four angles:

• With customers: advised Fortune 500 teams on when to process data in the cloud and when to act on it on the device (the edge).
• With the community: led the internal IoT community of more than 2,000 Microsoft employees, focused on IoT, Edge AI and Azure architectures. Its activities and events connected the field, the product teams and the content teams, and shaped the advice consultants gave their customers.
• With IoT experts: helped create the IoT workloads in the Azure Well-Architected Framework, then guided several customers through it, applying all five pillars, from reliability and security to cost, operations and performance.
• In the docs: helped introduce the cloud-connected and edge-connected patterns in Microsoft’s official documentation, giving teams that needed more intelligence at the edge a clear path to Edge AI.

The AI Strategy Advisory

If the assessment shows that some decisions belong on the device, and you’d like my help, here’s how I’d guide your team from roadmap to fleet. The report and the plan from the assessment are yours either way.

First months — roadmap:

• Device-to-cloud split for your shortlisted decisions
• Retrofit your existing hardware or redesign
• Security across your fleet, including secure software and model updates, software bills of materials (SBOMs) and vulnerability reporting under the CRA
• Pilot plan with success metrics

After the roadmap — advisory, for as long as it’s useful to you:

• Steer the pilot with your team and review results together
• Review vendor and hardware choices
• Guide the rollout from pilot to fleet
• Train your team to own the rollout