Our Mission
Tokowaka is Adobe's edge optimization platform for Large Language Model Optimization (LLMO). As AI-powered search and conversational interfaces become the primary way users discover information, websites need to ensure their content is fully accessible to AI crawlers. Tokowaka bridges this gap by serving prerendered, content-rich HTML to AI bots at the edge.
The name "Tokowaka" reflects our commitment to speed and agility. Operating at the CDN edge — the closest point to the requester — we intercept AI crawler traffic and transform JavaScript-heavy pages into fully rendered HTML that LLMs can parse and cite.
The Problem We Solve
Modern websites rely heavily on client-side JavaScript to render content. While this provides rich interactive experiences for human users, AI crawlers like GPTBot, ClaudeBot, and PerplexityBot often see empty or partial content. This leads to:
- Lost citations — LLMs can't cite content they can't see
- Reduced brand presence — competitors with crawlable content get cited instead
- Inaccurate summaries — partial content leads to incomplete or wrong AI-generated answers
- Missed traffic — users following AI recommendations end up on competitor sites
"If an AI can't read your content, it can't recommend your brand. Tokowaka makes sure every page is fully visible to every AI crawler."
How We Built It
We started by building bot detection at the edge. The tokowaka worker identifies AI crawler user agents (GPTBot, ClaudeBot, PerplexityBot, Googlebot, and others) and routes their requests through the prerender pipeline.
We integrated with a high-performance prerender service that executes JavaScript, waits for content to load, and returns fully rendered HTML. The prerendered content is cached at edge locations for subsequent requests.
We built the prerender-gains analyzer to quantify the impact. For each URL, we measure the word count before and after prerendering, calculating the content gain ratio. This data proves the value of edge optimization to stakeholders.
Integration with SpaceCat brought continuous monitoring — automated audits for LLM crawlability, brand presence tracking across AI platforms, and citation rate analysis. Teams can now see their LLMO metrics in the Elmo UI dashboard.
The Team
Tokowaka is built by the LLMO Optimizer at Edge team within Adobe's Experience Cloud organization. We work closely with the SpaceCat platform team, the AEM Edge Delivery team, and customer success teams to ensure every onboarded domain gets maximum value from edge optimization.
Our team spans engineering, data science, and product management. We operate with an autonomous, ship-fast mentality — continuously deploying improvements to the edge worker and expanding our domain coverage.
Get Started
Ready to onboard your domain? The process is straightforward:
- Onboard via SpaceCat:
@spacecat onboard site https://yourdomain.com - Set up domain comparison URLs in
prerender-gains/urls/ - Add your edge-dev API key to the environment config
- Run your first comparison:
./compare-domains.sh www-yourdomain - Review the Executive Summary and share with stakeholders