Why AI matters at this scale

Avista is a regulated utility providing electricity and natural gas to communities across the Pacific Northwest. Operating in a complex environment of aging infrastructure, climate volatility, and the energy transition, the company faces constant pressure to improve reliability, efficiency, and customer service while managing costs. For a mid-sized utility like Avista (1,001-5,000 employees), AI is not a futuristic concept but a pragmatic tool to tackle these core challenges. At this scale, the company generates vast operational data but may lack the internal resources of a tech giant. Strategic AI adoption allows Avista to punch above its weight, automating insights from data to make better capital and operational decisions, ultimately benefiting ratepayers and staying competitive in a changing energy landscape.

Concrete AI Opportunities with ROI

1. Predictive Grid Maintenance: Traditional maintenance is schedule-based or reactive. AI can analyze data from sensors, weather models, and historical failure rates to predict specific asset failures (e.g., a transformer overheating) weeks in advance. The ROI is clear: preventing a single major outage saves hundreds of thousands in restoration costs, avoids regulatory penalties, and dramatically improves customer satisfaction scores. For a utility of Avista's size, a 10-15% reduction in unplanned outage hours could translate to millions in annual savings and reliability incentives.

2. Dynamic Load and Renewable Forecasting: Integrating wind and solar power adds variability to the grid. AI models that fuse consumption patterns, weather forecasts, and even calendar events can predict local energy demand and renewable output with high accuracy. This allows Avista to optimize power purchases, reduce reliance on expensive peaker plants, and seamlessly balance the grid. The financial impact is direct: more efficient energy procurement and reduced congestion costs, protecting customer rates.

3. Intelligent Customer Engagement: AI-powered chatbots and analytics can personalize customer interactions. Bots can handle routine queries and outage reports 24/7, while predictive analytics can identify customers likely to struggle with bills and proactively offer payment plans or efficiency tips. This improves customer satisfaction (a key regulatory metric) while reducing call center volumes, allowing human agents to focus on complex, high-value interactions.

Deployment Risks for a Mid-Sized Utility

For a company in Avista's size band, AI deployment carries specific risks beyond technology. First, talent scarcity is acute; attracting and retaining data scientists and AI engineers is difficult and expensive, often requiring partnerships with vendors or consultants. Second, data silos are common; operational technology (OT) data from the grid and information technology (IT) data from business systems are often separate, requiring significant integration effort to create a unified AI-ready dataset. Third, the regulatory hurdle: any major investment must be justified in rate cases, and the benefits of AI projects must be clearly quantifiable to gain regulatory approval. Finally, change management within a traditionally engineering-focused culture can be slow; demonstrating quick, tangible wins from pilot projects is essential to build organizational buy-in for broader AI transformation.