AWS re:Invent 2025 -Scaling Global Energy Transformation: Kraken's AI-Powered Journey on AWS-IND3324
Source: Dev.to
Overview
AWS re:Invent 2025 – Scaling Global Energy Transformation: Kraken’s AI‑Powered Journey on AWS (IND3324)
Tyler (Kraken) presents the company’s AI philosophy for the energy and utilities sector. Kraken is a cloud‑native operating system for energy utilities built on AWS. It unifies customer care, billing, flexibility management, generation optimization, and field‑workforce tools under a single data model, enabling ~200 deployments daily.
Key AI capabilities
- Magic Ink – automated email responses.
- Customer Intent Dashboard – sentiment analysis.
- AI‑driven field scheduling – adds ~150 k site visits per year.
Security is enforced by redacting data before it reaches LLMs, and clients can choose from multiple LLM providers. The roadmap features Agent Studio, allowing energy specialists to manage autonomous AI‑bot teams.
Proven at scale with Octopus Energy (8 M customers), Kraken’s AI tools achieve an 89 % acceptance rate and boost both customer satisfaction and employee retention.
Introducing Kraken: An Operating System for the Energy Utility Industry
Why Kraken?
Kraken aims to transform the energy system by acting as an operating system for energy utilities worldwide. Much like AWS provides a platform for developers, Kraken provides a modern, unified platform for an industry that has long relied on siloed, legacy systems.
Four Fundamental Aspects
-
Customer Care & Billing Engine
- Consolidates Customer Information Systems (CIS), CRM, and meter‑data‑management into a single unified model.
-
Flexibility Management
- Controls distributed assets (smart thermostats, batteries, EVs) to balance the grid at the residential level.
-
Generation Optimization
- Optimizes power‑plant output and grid‑scale batteries to deliver value.
-
Field Workforce Tools
- Schedules and dispatches field crews, enabling efficient on‑site work.
These components must work together for modern utilities, and Kraken provides the first platform that does so end‑to‑end.
Built on AWS
- Cloud‑native: Designed from the ground up on AWS services.
- Singular Data Model: A purpose‑built energy data model eliminates traditional data silos, forming the foundation for AI.
- CI/CD at Scale: Supports ~200 deployments per day—far beyond the typical annual change cadence in utilities.
- Global Resilience: AWS’s worldwide infrastructure offers the reliability utilities require.
Kraken’s layered architecture (core, services, APIs, UI) runs on AWS.
AI Journey
Core AI Features
- Magic Ink – Generates automated email replies.
- Customer Intent Dashboard – Real‑time sentiment and intent detection.
- AI‑Driven Field Scheduling – Adds ~150 k additional site visits annually.
Security & LLM Flexibility
Data is redacted before being sent to large language models. Clients can select from multiple LLM providers, ensuring compliance and control.
Future Roadmap: Agent Studio
Agent Studio will enable agentic AI, allowing energy specialists to create and manage autonomous AI bot teams that can act on behalf of the utility.
Proven Scale
- Octopus Energy: 8 million customers.
- AI Acceptance Rate: 89 %
- Impact: Improves customer satisfaction and employee retention.
Additional Resources
- Full session video: AWS re:Invent 2025 – Kraken AI Journey (YouTube)
- Kraken website: https://www.krakenenergy.com