How dbt Labs Reduced Time Spent on Security Questionnaires by 50% With Conveyor

About

dbt Labs’ industry-standard platform enables analytics engineers to transform, test, and document data in cloud warehouses. Valued at $4.2 billion with backing from Snowflake, Databricks, and Salesforce Ventures, dbt is used by over 80,000 teams, including J.Crew, Nasdaq, and Conde Nast.

‍

Challenge

Scaling security reviews to support enterprise growth

With dbt Labs moving upmarket and closing enterprise deals, the volume and complexity of the security questionnaires they received surged.

Randy Hanooman, Senior Manager of Information Security at dbt Labs, leads the three-person team responsible for completing these requests. Questionnaires are just one part of the team’s responsibilities, but they’re a critical function that directly supports sales. Each one shapes customer trust and impacts deal velocity.

dbt Labs’ existing questionnaire tool suggested answers through database matching (searching a library for previous, related responses) but only yielded usable automated answers for about 35% of questions.

The extensive and time-consuming manual verification (especially for requests from third-party vendor portals) was quickly becoming unsustainable, not just because of growing volume, but because it was pulling the team away from audits, control testing, and other compliance work. Assessments often included hundreds of questions, meaning each could take anywhere from fifteen minutes to two hours, depending on complexity.

Randy knew that building and maintaining a complete, accessible knowledge library would be essential for speed and consistency. If his team could proactively identify and close information gaps, they could answer faster and keep deals moving, rather than having gaps only surface when a customer asked something that hadn’t yet been documented.

With 40 questionnaires coming in per month and sales goals increasing year over year, Randy needed a more scalable system to match the pace of growth. “We had the volume and the expertise to handle enterprise questionnaires,” Randy explains. “What we needed was a way to automate the repetitive work.”

Randy knew the only path forward involved AI, specifically a tool that could learn and improve over time, integrate directly with third-party portals, and proactively surface knowledge gaps.

dbt Labs was already using Conveyor’s Trust Center to share security documentation with prospects, and the partnership had been strong for years. When Conveyor introduced AI-powered questionnaire automation, Randy saw a way to expand that relationship.

‍

“Our goal is to focus on the things that can’t be automated, like control testing for audits and identifying security risks. But the questionnaire volume was pulling us away from that strategic work.”

‍

Solution

AI that transforms questionnaires into a scalable revenue engine

To begin, Randy’s team ran a trial processing real questionnaires through Conveyor’s AI. Initial first-pass accuracy started at 40–50%, but as the team built out the Knowledge Library and Conveyor’s AI learned from each approved response, accuracy climbed to 60-70% by the end of the trial, more than double what their previous tool achieved.

Conveyor works by analyzing each question, searching the knowledge graph (a structured database of interconnected documents, Q&A pairs, prior answers, and external sites/content), and generating a response. With 65% of answers ready to approve on first pass, the team only spends time on the 35% that need refinement, a fraction of the manual effort required before.

The system’s accuracy continues to improve over time, as Conveyor is designed to learn continuously. Every time Randy’s team approves or edits a response, that feedback strengthens the Knowledge Library, while the AI Librarian keeps content synced and current without manual maintenance.

Adoption was seamless, as the platform integrates with the tools the team already uses. Salesforce automates Trust Center access and helps the team determine whether to process a questionnaire by checking lead records. Slack enables auto-approval of access requests and lets the team get instant answers from the Knowledge Library. These integrations handle the tedious work beyond question answering itself. And for questionnaires that arrive through third-party vendor security platforms, Conveyor’s browser extension works directly within those portals.

dbt Labs has been an early adopter of Conveyor’s newer features, including the Knowledge Explorer, which proactively identifies gaps in their Knowledge Library. After evaluating dbt Labs’ Knowledge Library against the most common industry questions, the Explorer found that web accessibility was an area needing greater coverage. Randy’s team was able to proactively add the information before a customer even had to ask.

What once consumed 80+ hours a month now takes half that. The same three-person team handles the growing volume with capacity to spare for the audits, control testing, and vendor risk work that requires human judgment. And as sales goals grow, the automation will scale alongside them.

‍

“The more we automate, the more time we have to solve the hard problems. Conveyor frees up our team to focus on audits, vendor risk, and the work that actually strengthens our security posture.”

‍

Results

65% first-pass accuracy transforms security questionnaire workflow

What started as a Trust Center partnership has evolved into an integral part of dbt Labs’ Customer Trust program. The Security team now uses Conveyor daily for security questionnaire automation and monthly security metric reporting to tie their impact to revenue.

The impact is measurable:

• 65% first-pass accuracy on questionnaires
• 40%+ reduction in questionnaire time
• 5+ hours a week reclaimed for strategic work

Looking ahead, questionnaire volume will continue to grow alongside dbt Labs’ sales goals. As the Knowledge Library expands, accuracy will keep improving, freeing even more time for Randy’s team to focus on upgrading dbt Labs’ compliance program and evolving it to support cutting-edge AI and software development.

‍