How AccuKnox built its content engine in Magi
AccuKnox sells cloud-native security into categories that each have their own buyer. The team mapped those buyers into Magi first, and the drafts stopped reading like a stranger wrote them.
- 11
- markets mapped, each with its own ICP, competitor set and influencers
- 33
- personas built with the pain points and KPIs each buyer is judged on
- $2M
- influenced pipeline across inbound and partner channels
- Industry
- Cloud-native security, Zero Trust CNAPP
- Company size
- Venture-backed, around 200 people across four continents
- Marketing team
- Seven contributors, led by a founding product marketing and partnerships lead
- Use cases
- Blogs, LinkedIn, email, one-pagers, webinars, landing pages, research
- Engagement
- Magi platform, self-serve
Different buyers in every category, and AI drafts that pitched them all the same.
- AccuKnox sells into CNAPP, ASPM, CSPM, CWPP, KSPM, AI security, API security, DSPM, compliance, 5G security and MSSP partnerships, each with its own buyers and compliance needs. The backlog outgrew any one writer, and research ate execution time.
Built the knowledge architecture first, then generated.
- Each market got its own ICP, personas, competitors and sources. Publishing rules and SEO frontmatter went into BrandOS, so every draft starts from one standard.
Buyer-specific content shipping across every market at once.
- Content batches ship in hours, not days. Every idea arrives tied to a campaign, a source and the buyer it is for.
From single-market focus to parallel multi-market execution.
Each market needs localized messaging for its own buyers, industries and compliance rules, all at once. The team could only work one market at a time, and without buyer context the AI pitched them all the same. A CNAPP buyer wants less cloud risk. A platform engineer wants fewer runtime alerts.
One market, one ICP, three personas: the user running the tooling, from DevSecOps to platform engineers; the buyer with budget, from CISOs to AI governance leads; and the channel partner. The team built it by hand over weeks of working sessions, before generating anything.
Each market in Magi is a campaign anchor. Open AI-SPM and its ICP, buying cycle, competitors, influencers and personas sit in one place, so AI-SPM drafts are written for AI-SPM buyers.
A competitive map for every category, monitored in one view.
Without a competitive map, the AI writes like it has never seen the market. Broad platforms crowd some categories and specialists hold others, so positioning shifts with each.
Each market got two broad platforms and two niche competitors, each with site, LinkedIn, blog and a positioning summary, plus a short list of influencers tracked for signal.
Magi's research reports pull competitor posts, blogs and CVE coverage into one view per market. Company and persona intel that took hours per campaign is now on demand.
A publishing standard enforced at generation time.
Metadata is the difference between readable and shippable. The standard has fourteen mandatory components, and missing one sent a draft back.
Every blog carries required frontmatter: title, excerpt, slug, author, meta title and description, TLDR, keywords, sticky CTA and FAQs. Landing pages map to parent and child pages, so new ones fit the site structure.
BrandOS holds frontmatter, tone, category rules and required links, set globally or per content type. Fourteen templates carry the standard, and drafts arrive close to publishable.
Every source turned into ideas, tagged to the campaign it serves.
AccuKnox's best explainers lived in sales calls, POCs, advisories and internal docs. A library nobody maps to a campaign is dead storage.
The team loaded FAQs, llms.txt, whitepapers, CVE pages, press releases, docs, playbooks, decks, POC summaries, call transcripts and objection logs, and mapped each to a campaign. From there the loop runs: collect, classify, map, generate.
Collections group sources by campaign. Each idea opens as a brief with a TLDR, the source excerpt behind it and a button to draft.
Content as a queue to clear, not a blank page to fill.
Content teams work in deadlines, campaign timing and channel mix, not isolated prompts. A tool that cannot hold those stays a demo.
Blogs, landing pages, emails and one-pagers share one calendar, each tagged by campaign, format, owner and status. Magi takes the first cut from a brief, and the team fine-tunes at the end.
Each draft is grounded in a specific source, campaign and ICP, with brand rules applied. The result: sharper buyer framing, stronger SEO structure and phrasing closer to how the team talks.
Six months on Magi
Apr–Sep 2026
- 4–6 hrs
- to produce a content batch, down from 3–5 days
- 2,779
- research runs
- 1,506
- knowledge artifacts built
- 1,673
- ideas generated
- 91
- ideas promoted to production
- 73
- drafts from 77 outlines
- 87
- documents built from templates
- 100%
- template adoption
Next: every market, in every format.
AccuKnox is extending the setup from blogs to landing pages, email sequences and one-pagers across every market, using the page map already in place. Each new source becomes ideas for the campaigns it maps to, so the library compounds. The model was never the bottleneck. The knowledge architecture was.
Days of coordination, now one workflow
"Magi transformed how we scale content across 11 markets. What used to take days of coordination now happens in a single workflow, from ICP-mapped blogs to SEO-ready copy, all on-brand. It's now the backbone of our content team."
Atharva Shah
Technical Content Lead & DevRel, AccuKnox
Other customer stories
Selling into more markets than your team can write for?
AccuKnox mapped its buyers into Magi and turned them into a content queue. See what yours would look like.