Client Overview
The client is a leading provider of renewable energy solutions, delivering smart energy services powered by IoT technologies. With a strong focus on digital innovation, the company collects energy data from a wide range of sources to provide advanced analytics, optimise operations, and enhance customer engagement. The client’s commitment to intelligent automation led to its adoption of Generative AI to improve data accessibility and user experience.

Challenge
The client set out to revolutionise how energy data is accessed and understood. With over 100 TB of IoT-generated data, they needed more than just scalable storage and processing, they required a way for users to explore their energy usage intuitively. Key challenges included:
Static and Technical Data Access
Energy data was complex and technical, offering limited accessibility to business users or customers without engineering expertise.
Need for AI-Driven Engagement
The client aimed to go beyond dashboards and enable interactive, GenAI-powered experiences for real-time energy insights.
Scalability and Data Management
Managing more than 100 TB of IoT-generated energy data required a robust cloud-native architecture capable of scaling efficiently.
Lack of Unified Observability
Without full observability, it was difficult to monitor infrastructure performance or ensure timely issue resolution.
Solution
Empowering energy data with scalable infrastructure and AI-driven customer interaction.
Bion Consulting delivered a Generative AI-enabled data platform on AWS, integrating LLMs to bridge the gap between complex energy data and everyday user interaction.
Generative AI Enablement
- LLM Integration via AWS Bedrock: Trained and fine-tuned large language models to understand and summarise energy data.
- Conversational Interfaces: Built customer-facing interfaces enabling users to ask natural language questions about their energy usage and receive AI-generated insights instantly.
Scalable Big Data Infrastructure
- IoT Data Pipeline Setup: Deployed streaming and batch pipelines for continuous ingestion of IoT data using AWS Kinesis and Glue.
- High-Volume Data Storage: Architected scalable storage and analytics with AWS S3, Athena, Redshift, and Lambda to support current and future data growth.
Real-Time Monitoring and Visualisation
- Observability Tools: Integrated New Relic for infrastructure health monitoring and performance tracking.
- AI-Supported Dashboards: Developed Amazon QuickSight dashboards enhanced by AI-generated context, improving usability and decision-making.
Results

Conversational AI for Energy Data
Enabled customers to query data via natural language, providing real-time, human-like responses and driving stronger engagement.

Enhanced Decision-Making
LLM-backed insights helped internal teams and customers make smarter, faster energy decisions, improving visibility and responsiveness.

Scalable Data Ecosystem
The AWS-powered infrastructure supports over 100 TB of IoT data, ensuring scalability, performance, and future growth.

Unified Observability and Control
End-to-end monitoring improved system reliability and enabled proactive optimisation through data-driven alerts and insights.
Technology Stack
To build a scalable and AI-enabled energy platform, the following technologies were implemented:- AI/ML: AWS Bedrock, Amazon SageMaker
- Big Data & Streaming: AWS Kinesis, AWS Glue
- Cloud & Storage: AWS S3, AWS Lambda, Amazon Redshift, Amazon Athena
- IoT Integration: AWS IoT Core
- Visualisation & Monitoring: Amazon QuickSight, CloudWatch, Grafana
- Infrastructure-as-Code: Terraform
