How Novara scaled from a handful of RFPs to 75–100 a year with the same five solutions engineers

Novara's SE team uses Inventive AI to answer product RFPs and security questionnaires from live product documentation.

Checkmark icon inside a peach-colored circle on a white background.
75–100 RFPs a Year, No New Hires

The same five solutions engineers cover the jump in volume.

Checkmark icon inside a peach-colored circle on a white background.
Every Answer Cited Back to Its Source

Outdated answers trace back to the exact document to fix.

Checkmark icon inside a peach-colored circle on a white background.
4 Tools, Built Right In

Runs on SharePoint, Salesforce, Slack and SSO.

Novara and Inventive AI case study cover

For Novara, the safety and operational risk software company behind the Flex platform, enterprise RFPs and IT security questionnaires were suddenly arriving far more often. Its five-person solutions engineering team was used to a couple of RFPs a year, not 75-100. With Inventive AI, the team drafts every answer from live product documentation instead of a folder of old responses. This case study is based on conversations with Dita Amtey and Arlo Amis, Solutions Engineers, and Nick Connor, VP of Technology and CISO, at Novara.

Impact at a Glance

Novara impact at a glance: Scaling Without Adding Headcount

About Novara

Novara builds Flex, a safety and operational risk management platform for industries such as construction, manufacturing, utilities, and oil and gas. A specialized five-person solutions engineering team answers its product RFPs and IT security questionnaires.

The challenge: trapped in a "folder of folders" with stale data

Novara's pre-sales team knew a wave of enterprise interest was coming, but its knowledge system wasn't built for it. Past responses were scattered across Word documents, PDFs and spreadsheets in one unsearchable drive.

"It's literally just a folder with folders within it of all of our different prospects that have reached out to us over the past few years. Everything from Word to PDF to Excel documents."
— Dita Amtey, Solutions Engineer, Novara

The team expected volume to jump to 75–100 proposals a year. Handling that by hand, especially repetitive IT security questionnaires, wasn't realistic for five people. And an answer pulled from a year-old RFP was likely to describe features that had since changed.

Leadership knew it needed AI to scale, but it had to work for enterprise pre-sales.

"Ethan just messaged me this morning and was like, 'How good is our enterprise ChatGPT in filling out RFPs?' So I told him we actually just bought a new tool."
— Arlo Amis, Solutions Engineer, Novara

The solution: a Slack-integrated proposal engine

After a competitive proof of concept, Novara chose Inventive AI for RFP automation for solutions engineers, making it the team’s central proposal engine.

When a security questionnaire arrives, the SE team drops it into Inventive AI, which drafts every answer from a curated SharePoint folder of live product documentation. Projects launch from Salesforce and access runs through SSO via Inventive's integrations, so the team kept its existing workflow.

For the wider sales team, the turning point was the Slack bot, which answers one-off customer questions on the spot.

"I just asked you, what is our uptime SLA? And we commit to 99.5% monthly uptime. So it's actually pulling from that source document because it's referencing how it's calculated."
— Nick Connor, VP of Technology and CISO, Novara

How Novara's AI drafts beat what SEs had time to write

Inventive AI's drafts were often more thorough than the answers the SEs had time to write by hand.

"I've answered it a lot with just 'yes.' So this is more detail than I give a lot of the time."
— Arlo Amis, Solutions Engineer, Novara

It also fixed the stale data problem. Inventive AI gives answers with source citations, so when the team spots an outdated response, it can trace it to the SharePoint file, update that file once and keep the fix for every future proposal.

Why Novara's solutions engineers trust where every answer comes from

"As we identify responses that are outdated, we can source that response back to the document it got it from, so that we know what needs to be removed or updated."
— Arlo Amis, Solutions Engineer, Novara

Answer security questionnaires from live product docs with Inventive AI

Inventive AI helps solutions engineering teams answer RFPs and security questionnaires with accurate, cited answers from approved content.

Book a demo to see how Inventive AI handles questionnaires like Novara's.

Frequently Asked Questions

Is enterprise ChatGPT enough for filling out RFPs?

It can answer single questions, but it doesn't work through whole questionnaires, cite where each answer came from or keep answers tied to current product documentation. Purpose-built RFP software does those things, which is why Novara bought a dedicated tool when leadership asked about using enterprise ChatGPT.

How do solutions engineers handle repetitive IT security questionnaires?

They keep one curated folder of current security and product documentation and let AI draft each questionnaire from it. SEs then review only the answers that are new, unclear or low confidence, instead of retyping the same responses every time.

What should a source folder for AI RFP answers contain?

Current product documentation, security policies, certifications, SLAs and approved boilerplate, with old versions removed. The AI can only be as current as the documents it reads, so the folder needs an owner who updates it whenever the product changes.

Can Inventive AI automate RFPs for a solutions engineering team?

Yes. Novara’s five-person SE team uses Inventive AI to draft product RFPs and IT security questionnaires from a curated SharePoint folder of live product documentation.

Does AI RFP software show where an answer came from?

Good ones do. Inventive AI cites the source document for every answer, so reviewers can check a claim like an uptime SLA against the exact file it came from.

How do you keep AI RFP answers from going stale as the product changes?

Trace each outdated answer back to its source document and fix it there once, so every future draft picks up the correction.