Uploaded February 2026 | Updated September 2026, 2 weeks ago
Can serverless architecture keep the lights on in London? đź’ˇ
As the world shifts to unpredictable renewable energy like wind and solar, the power grid faces a massive stability challenge. In this deep dive, Kevin Bowman (Engineering Lead at Kraken/Octopus Energy) reveals how they use AWS Lambda, EventBridge, and IoT to manage gigawatt-scale batteries and half a million electric vehicles in real-time.
Learn how Kraken avoids "reinventing the wheel" by leveraging managed services to solve complex physics problems, manage high-volume telemetry, and maintain 24-hour autonomy even if the cloud goes offline. Kevin breaks down their "Trigger/Action" Lambda pattern, their approach to event-driven microservices, and why your architecture must account for the "blast radius" of cloud failures.
⏱️ Video Timestamps (For Navigation)
0:00 - Why London doesn't go dark when the wind stops
2:15 - The Physics Problem: AC Frequency & Grid Stability
4:50 - Batteries vs. Inverters: The need for an "Autopilot"
7:30 - How Energy Markets & Settlement Windows work
10:15 - Real-world Incident: The Jan 30th Grid Trip
12:45 - The Optimizer: Using SAT Solvers for 24-hour planning
15:10 - Why we chose AWS Serverless (Lambda & DynamoDB)
18:00 - Deep Dive: The Trigger/Action Lambda Pattern
21:30 - Handling 1,000s of data points with Redis & InfluxDB
24:45 - The Edge: Industrial PCs and IoT Hardware
28:20 - Scaling to 500,000 EVs (Virtual Power Plants)
31:40 - Q&A: Reliability, Incidents, and Vendor Lock-in
đź”— Transcript available on InfoQ: bit.ly/4kiHQrb
#SoftwareArchitecture #Serverless #AWSLambda #Energy #EngineeringLeadership #GreenTech #infoq
Can serverless architecture keep the lights on in London? đź’ˇ
As the world shifts to unpredictable renewable energy like wind and solar, the power grid faces a massive stability challenge. In this deep dive, Kevin Bowman (Engineering Lead at Kraken/Octopus Energy) reveals how they use AWS Lambda, EventBridge, and IoT to manage gigawatt-scale batteries and half a million electric vehicles in real-time.
Learn how Kraken avoids "reinventing the wheel" by leveraging managed services to solve complex physics problems, manage high-volume telemetry, and maintain 24-hour autonomy even if the cloud goes offline. Kevin breaks down their "Trigger/Action" Lambda pattern, their approach to event-driven microservices, and why your architecture must account for the "blast radius" of cloud failures.
⏱️ Video Timestamps (For Navigation)
0:00 - Why London doesn't go dark when the wind stops
2:15 - The Physics Problem: AC Frequency & Grid Stability
4:50 - Batteries vs. Inverters: The need for an "Autopilot"
7:30 - How Energy Markets & Settlement Windows work
10:15 - Real-world Incident: The Jan 30th Grid Trip
12:45 - The Optimizer: Using SAT Solvers for 24-hour planning
15:10 - Why we chose AWS Serverless (Lambda & DynamoDB)
18:00 - Deep Dive: The Trigger/Action Lambda Pattern
21:30 - Handling 1,000s of data points with Redis & InfluxDB
24:45 - The Edge: Industrial PCs and IoT Hardware
28:20 - Scaling to 500,000 EVs (Virtual Power Plants)
31:40 - Q&A: Reliability, Incidents, and Vendor Lock-in
đź”— Transcript available on InfoQ: bit.ly/4kiHQrb
#SoftwareArchitecture #Serverless #AWSLambda #Energy #EngineeringLeadership #GreenTech #infoq
