Elasticsearch was a nightmare to maintain and didn't work well for our use case. With Ducky, we shipped a brand new feature in two weeks that our customers immediately loved. The search just works , and it keeps getting better.
Ray Rhodes, Chief Product Officer and Co-founder, UserEvidence
UserEvidence helps companies capture and leverage social proof at scale. Their platform collects customer feedback through surveys and organizes it into searchable research libraries where sales teams, marketers, and product managers can find testimonial infographics, charts, statistics, case study documents, and video content - whatever asset they need to support their work.
When UserEvidence's customers needed faster, more intuitive search across their growing content libraries, the team knew their existing Elasticsearch setup wasn't cutting it. They needed something better and faster.
Challenge
Solution
Implemented Ducky AI’s fully managed search infrastructure
Multi-modal search across text, images, and structured data
Automated chunking and reranking for optimal results
Advanced metadata filtering for precise queries
Results
Launched new search feature in just two weeks
Unlocked brand new report generation capabilities previously impossible
Dramatically reduced time for sales teams to find relevant materials
Improved customer satisfaction with faster, more accurate search results
The Challenge
UserEvidence built their platform to solve a core problem: companies struggle to collect, organize, and use customer feedback effectively. Their libraries aggregate thousands of survey responses, quotes, and visual assets — all the social proof go-to-market teams need.
But that content is useless if people can’t find it.
“Our customers were collecting incredible insights, but sales teams spent too much time hunting for the right quote or statistic” says Ray. “We knew search had to be the foundation of everything we built.”
The team initially chose Elasticsearch for flexibility, but it quickly became a burden.
“Elasticsearch is powerful, but it needs constant tuning. Every time we wanted better results, we had to dig into configuration files” Ray explains. “Even then, it couldn’t handle the nuance of searching across quotes, data, and images.”
Emerson Loustau, Senior Software Engineer, puts it bluntly: “We still have Elasticsearch in our stack, and not a week goes by without an issue or something breaking.”
The complexity slowed product development and limited what UserEvidence could deliver. Features like generating reports from search results stayed out of reach.
“We couldn’t ship what customers needed because we were too busy keeping search infrastructure alive.”
We love the UserEvidence team
The Solution
UserEvidence needed a search solution that would let them focus on building features instead of maintaining infrastructure. They found it in Ducky AI. UserEvidence evaluated several options, but Ducky’s speed, simplicity, and developer-first approach won them over.
“Three things made Ducky the obvious choice” Ray says. ‘First, it’s incredibly fast with zero maintenance. Second, multi-modal search works seamlessly across all our content. Third, it was so simple to implement — we went from decision to production in two weeks.”
Unlike Elasticsearch, which required building and maintaining complex pipelines, Ducky runs everything automatically. Its multi-modal search lets UserEvidence find quotes, survey data, and visuals in one place instead of treating them as separate problems.
“Our engineers loved working with Ducky. The API is clean, the docs are excellent, and everything just works” Ray adds.
Within two weeks, UserEvidence had launched a powerful new search experience.
The Results
The impact was immediate. Sales teams could find exactly what they needed in seconds instead of minutes, filtering by customer segment, product feature, or date range to uncover the perfect quote or data point. Every search felt effortless, letting them focus on their goals instead of digging for proof.
“The feedback was incredible. Our customers told us search went from frustrating to delightful” says Ray. “Results were more relevant because Ducky automatically handles the hard parts — splitting documents intelligently and ranking results in multiple stages. We didn’t have to tune anything. But the bigger win was what this unlocked for our product roadmap.”
With reliable search in place, UserEvidence quickly shipped report generation capabilities that pull relevant content and synthesize it into customer-facing materials.
“Ducky’s RAG support means we can do more than just return search results. We can feed those results to LLMs and generate polished reports automatically” Ray explains. ‘This was impossible with our old setup. Now it’s a core product feature.”
Marketing teams now generate case study materials in minutes. Sales teams pull together customized pitch decks. Product managers compile feedback reports, all without manual copying and pasting.
Technical wins
What is Next
UserEvidence continues expanding what’s possible with Ducky-powered search, exploring more sophisticated report generation, automated insight extraction, and personalized recommendations.
“Effective search and discovery isn’t just a feature for us, it’s the foundation of everything we build”, Ray reflects. “Ducky gave us that foundation, and now we’re building things we never thought possible.”
