Quilt review: pricing, features and alternatives (2026)

The verdict up front

Quilt was a sharp little product with a clear audience, and you probably can’t buy it the way it used to be sold. That’s the whole review in two sentences. The rest explains why it was worth a look, what changed, and what to ask before you spend a minute on a demo.

Quilt built AI assistants for solutions engineers. The flagship drafted answers to RFPs and security questionnaires from your own documents, and a second assistant answered the “hey, do we support SAML?” pings that reps fire into Slack all day. Sequoia led a $2.5 million seed round in March 2024, according to TechCrunch. Then the quilt.app homepage announced that Quilt had merged with Rox, a revenue-agent platform, and told new buyers to try a Rox autofill product instead. Existing users, the banner said, could keep using Quilt as normal.

When we went to check the live site in October 2026, quilt.app sat in a redirect loop and never loaded. The Rox autofill page it pointed to now redirects to Rox’s general platform page. Sequoia’s portfolio page lists Quilt as founded in 2023 and acquired in 2025.

So here’s the honest call. If you already run Quilt, ask Rox for a written roadmap and an export plan. If you’re shopping fresh, treat Quilt as a reference point for what a solutions-team tool can look like, and put your evaluation time into a vendor that’s still selling it as a standalone product.

What it actually does

Quilt split its product into assistants, each aimed at a specific SE chore.

The Questionnaire Assistant took an RFP, RFI, security questionnaire or DDQ and produced a first pass from your connected sources. Quilt’s own product page claimed a 95% completion rate on first pass. That’s a vendor number, so read it as Quilt’s own estimate, not a benchmark. The page also described a spreadsheet-style workspace for assigning questions and a Chrome extension that pulled questions out of vendor portals and wrote answers back.

The Knowledge Assistant was the quieter, more interesting half. It sat in Slack and Chrome and answered internal questions from reps, with ranked sources under each answer. If you’ve ever been the SE who answers the same SSO question in four different channels before lunch, you get the appeal. Quilt’s seed announcement framed it as managing your “#ask” channels around the clock.

A third piece, a meeting assistant for call prep, was listed as coming soon on the product menu. The Enterprise plan description mentioned an “intelligent meeting assistant,” but the product page itself never left the coming-soon label in the captures we read.

How drafting and the knowledge base work

Quilt’s approach was retrieval over your live sources rather than a hand-curated Q&A library. Its security FAQ is blunt about this: Quilt said it did not store a repository of questions and answers, and instead indexed the documents you shared with it and built a search index around them. Answers were generated on demand by sending the relevant snippets to OpenAI, alongside what the FAQ called a proprietary mix of open source and hosted models.

That design has a real upside. You don’t spend Friday afternoons pruning stale library entries, because there’s no library to prune. Update the policy doc in Google Drive and the next answer reflects it.

The downside shows up in review. Later product pages added an “Answer Bank” where you could correct a wrong answer in conversation so it didn’t repeat. That’s a sensible patch, but it means your source of truth is split between the documents and the corrections. In a demo, you’d want to see how those two interact when they disagree.

Connectors depended on plan. The last pricing page we could read listed Google Drive and web pages on Free, Slack on Team, and Notion, Confluence, Zoom, Chorus and Gong on Enterprise. The knowledge page claimed more than 20 integrations in total. G2’s vendor-supplied profile repeats Google Suite, Confluence, Notion and Zoom.

Source attribution was present but not the headline. The Knowledge Assistant page promises ranked sources so users can trust answers, and one 2024 G2 reviewer called the in-app citations helpful. That’s capable, not category-leading.

Security and compliance posture

Quilt’s homepage said it was fully SOC 2 Type 2 compliant and linked visitors to its report. It also said Quilt only accessed documents you uploaded or shared, and that it never fine-tuned shared models on customer data. Each customer’s data was kept isolated. Encryption was listed as TLS 1.2 or 1.3 in transit and AES-256 at rest.

The security FAQ adds detail a security reviewer would want. Data lived on Google Cloud, replicated across several North American regions. Document snippets went to OpenAI only when needed to answer a question. Deletion requests were handled by email within 30 days. Older FAQ answers said users authenticated through a Google account, while the pricing page put SSO and SCIM on the Enterprise plan.

None of this tells you what Rox does with the data today. Ownership changed, so the subprocessor list and the SOC 2 scope may have changed with it. Ask for the current report, not the old one.

Pricing and how to buy it

The last captured pricing page (August 2025) had three tiers. Free was $0 forever for 3 members and 20 questionnaires a month, with the Google Drive agent and the Chrome extension. Team added unlimited seats and questionnaires plus the Slack agent, with the price behind a demo. Enterprise added the meeting assistant, the broader connector list, SSO with SCIM, and a dedicated CSM, again with no listed price.

That free tier was genuinely useful for a small SE team that wanted to try it on a live questionnaire without a procurement cycle. It’s also the part we’d trust least today. With quilt.app offline and new buyers redirected to Rox, there’s no current page that confirms any of these plans still exist.

Where it falls short

The availability question dwarfs everything else. A tool you can’t sign up for, from a site that doesn’t load, is a hard sell no matter how good the drafting was.

Beyond that, the sourced weak spots are fairly typical for a young product:

• Team and Enterprise prices were never published, and the connectors most SE teams want, such as Confluence and Gong, sat on the top tier.
• G2 reviewers in 2024 said the review interface got awkward on large sheets with several collaborators, and one described formatting and edit history as fairly inflexible.
• Several 2024 G2 reviewers asked for better handling of PDF and Word questionnaires, and one noted that enterprise-grade reporting didn’t exist yet.

To be fair, the same reviews were warm overall. G2 showed a 4.9 out of 5 average across 16 reviews in the March 2025 capture we read, with time savings and ease of use as the recurring praise.

Who should shortlist it and who should skip it

Shortlist it if you’re an existing Quilt customer deciding whether to stay. In that case the real evaluation is of Rox, and your questions are about the roadmap and how you’d migrate your data out.

Skip it if you’re starting a new search. The product’s best idea, one knowledge layer that serves both questionnaires and rep questions, is available from vendors that are still actively selling and publishing current pricing.

How it sits next to its peers

Arphie is the closest match on philosophy: drafts from live-connected sources instead of a hand-kept library, aimed at presales and SE teams. Arphie has a broader published connector list and a current product page, and it shows confidence scores on answers. Quilt’s edge was the free tier, which Arphie doesn’t offer, since Arphie is quote-only.

Tribble covers the same two jobs Quilt did: questionnaire responses plus a sales answer layer inside Slack and Teams. Tribble is the more enterprise-shaped option, with integrations into CRMs like Salesforce plus call recorders like Gong. Quilt was the lighter, cheaper-to-start choice, though Tribble publishes no prices either.

1up is probably the best replacement for anyone who liked Quilt’s free plan. It has a free-forever tier and flat monthly paid plans, plus a Slack and Teams Q&A bot for sales enablement. Its catch is that questionnaire automations are capped by plan, which Quilt’s Team tier didn’t do.

Loopio is the opposite design: a mature, tagged answer library with multi-step review. If your problem is governance across a large proposal team, Loopio has years of workflow depth that Quilt never reached. If your problem is SE time spent on repeat questions, Quilt’s retrieval-first approach was the lighter lift.

What to test in a demo

If you do end up talking to Rox about Quilt, bring this list:

1. Ask, in writing, whether Quilt is still sold as a standalone product and for how long existing contracts will be honored.
2. Request the current SOC 2 Type 2 report and confirm it covers the Quilt product under Rox ownership.
3. Get the current subprocessor list, and confirm whether OpenAI is still the model provider and whether data stays in North American Google Cloud regions.
4. Ask how you export your Answer Bank corrections and project history if you leave.
5. Run one of your real security questionnaires through it, ideally a PDF or Word file, since reviewers flagged those formats.
6. Load a 200-row sheet with three reviewers working at once and see whether the interface holds up.
7. Check what reporting exists for completed projects, since reviewers called it thin.
8. Confirm which connectors your plan includes, and whether SSO still requires the top tier.

Our take: Quilt was a thoughtful tool for SE teams, and its retrieval-first design aged well. The company behind it has moved on. Your shortlist should too.