Intelligent Search at Scale
Transforming Digital Discovery Through Human-Centered AI
In an age of information overload, searching for spiritual guidance shouldn't feel like a test of faith. Yet for millions of church members, pastors, and staff across thousands of congregations, finding the right resources, whether it's a sacred text, a theological teaching, or details about a ministry, was becoming increasingly frustrating.
When 61% of website visitors abandon a site within five seconds if they can't find what they're looking for, and 75% of new ELCA site visitors were actively searching for content, the stakes were clear: the church's digital presence needed to meet people where they are, or risk losing the connection entirely.
The Full Scoop
Challenge
Across thousands of ELCA congregations, members, pastors, and staff struggled to find faith resources due to fragmented search experiences and disconnected sites. Navigation scored just 2.87/5, with 75% of new visitors actively searching and 13% going straight to the search box—yet keyword search consistently fell short.
The challenge wasn’t only technical. Faith-based queries are deeply personal and require responses grounded in theology, not just accuracy. Inconsistent results and manual processes created frustration for users and staff alike. The stakes were high: in a digital world where users form opinions in seconds, poor search experiences threatened the church’s ability to connect, serve, and fulfill its mission. ELCA needed more than a new tool—it needed a scalable, human-centered framework for spiritual discovery.
Approach
nvisia partnered with ELCA to redefine search through human-centered design, theological integrity, and technical scalability. Instead of rushing to implementation, the team validated both architecture and readiness through parallel proofs of concept—testing contextual understanding, prompt security, and performance at scale.
The solution replaced keyword search with semantic, AI-powered discovery built on Azure AI Foundry and a React front end. Multiple safeguards—including Azure content filtering, evaluators for accuracy, and adversarial testing—ensured reliability and trust.
Equally vital was the human side. Interviews with pastors, staff, and members surfaced pain points across distinct user journeys like Find a Congregation and Give to ELCA. Iterative prototype testing and faith-informed prompts kept every design decision aligned with pastoral principles.
This blend of advanced architecture and theological grounding allowed ELCA to modernize responsibly—balancing innovation with mission and ensuring the system would perform at scale while honoring the faith it serves.
Outcomes
ELCA’s new intelligent search transformed digital discovery from a frustrating task into a trusted, intuitive experience. Users now engage naturally with AI-powered search that understands context, honors theology, and connects seekers across four key journeys—Give to ELCA, Find a Minister, Find a Congregation, and Grow in Knowledge and Faith.
With 40,000+ documents unified across six sites, rapid response times, and 24/7 availability, the system delivers both performance and reliability. Theological guardrails and continuous evaluation preserve accuracy and trust.
Operationally, discoverability has improved across the board—pastors locate guidance faster, congregations access materials more easily, and seekers spend less time searching and more time learning. Support tickets have declined, while user sentiment has improved.
Most importantly, the project has reshaped ELCA’s digital culture, proving that AI can enhance—not replace—human connection when grounded in empathy, inclusivity, and faith.
Get in touch
Is your organization struggling to help users find the information they need across complex, multi-site digital ecosystems? We can help transform your search experience - let's talk.
