AI is already reshaping the professional learning market. Enterprise expectations are changing. Learner behavior is changing. That part isn't really in question anymore. What's less settled is what it does to BenchPrep's internal operations, and who within the company gets to act on it. How we change internally would determine how quickly we can unlock value for our customers.
We first framed that as a narrower, more familiar question: what should BenchPrep's internal AI layer be?
Our data already lives in Snowflake. Hence, Snowflake Cortex was the obvious first candidate. Cortex offers our team a direct path to query it. But that framing was too small. Querying data is one job. The real question was whether every team at BenchPrep, not just engineers and analysts, could solve real business problems by building tools, automating workflows, and testing ideas with AI, not just asking it questions.
That's the bigger bet. Synthetiq is how we're making it. We are using Synthetiq’s no-code environment for building internal apps, AI agents/workflow automation, natural-language data access, and more.
From a Pilot to a Company-Wide Capability
What began as a decision about internal AI quickly expanded. The scope grew fast. Teams across BenchPrep now use Synthetiq to:
- Turn raw data into an answer in hours instead of waiting on a data team's queue
- Go from idea to working prototype before the next customer meeting, not the next sprint
- Build the internal tool nobody had bandwidth to build
- Cut hours of manual, repetitive work down to minutes
The biggest unlock has been accessibility. Internal teams no longer need to be engineers to build a useful internal tool or data analysts to begin investigating a business question. The people closest to a problem can increasingly create the solution themselves.
For example, the AI Usage Dashboard gives customer success and product teams a per-customer view of how learners are using our AI capabilities, specifically the AI Assistant tool. It helps track usage volume, response quality, and engagement in one place. It turns raw interaction data into actionable insights, helping us prove value to customers, spot adoption gaps, and improve the product experience.
Another team used Synthetiq to reveal how learner engagement within a specific product translated into real score gains. They compared each student's initial diagnostic with their later benchmark exam across features such as custom practice, lessons, and flashcards. It gives our customers clear, evidence-backed proof of which study behaviors drive the biggest improvements, so they can double down on what works and guide struggling learners toward the activities most likely to lift their scores.
Aside from surfacing insightful data, Synthetiq has helped teams cut time on routine workflows by up to 70%.
Take our Release Notes generator: it pulls tickets flagged for release from Jira and translates engineering and product jargon into release notes customers can actually understand. Trained on our past release notes, it highlights the value of each update and how to activate it, without getting into technical detail.
Previously, each ticket had to be manually fed into an LLM to draft copy, taking our team 3-4 hours per release. Now, one link to the sprint board pulls and writes all tickets in bulk, cutting that down to about an hour, up to 70% of the time previously spent.
What We're Learning
Becoming AI-first isn't about rolling out tools. It's about value-driven adoption. It is about expanding who can build, experiment, and change how work gets done.
That's not a hard sell at BenchPrep. We're a learning company. Our mission is helping people build new skills. Asking our own teams to do the same wasn't a culture shift. It's the same culture, pointed inward.
Synthetiq doesn't replace engineering, data expertise, governance, or judgment. It changes where those resources are spent. Less resources on gatekeeping requests and more on the problems that actually need them. This shift enabled every employee to act on their own ideas.
We're early in this, but we know where it's headed: every BenchPrep employee automating a workflow, building a solution, testing an idea, or finding an insight — on their own, without waiting in line for someone else to do it.
We initially went looking for an internal AI solution. What we found was a platform that could help us rethink how the entire company works.
For a learning company, that feels exactly right.
