Conteply: Google Docs as an AI-orchestrated publishing surface

Conteply uses AI to turn Google Docs into a publishing platform. Non-technical marketing and agency teams can then write, approve, and ship website updates without CMS, dev tickets, or CI/CD cycles.

Business сhallenge

A marketing team working across multiple CMS environments and legacy stacks needed a faster way to ship website content updates. Even minor copy edits kept turning into engineering tickets and CI/CD cycles. This “publishing tax” created:

• Time-to-publish inflation. Routine text edits that should take minutes became multi-day release tickets.
• Developer churn. Engineers spent time on mechanical copy updates instead of product work, clogging queues and creating friction between teams.
• Content drift and compliance risk. Approved documents could diverge from live pages at any point in the cycle, creating brand inconsistency and potential legal exposure.

The team needed a practical solution where marketing could own content and update website copy independently, without CMS overhead or deployment waits, while keeping the familiar tools, like Google Docs.  

Solution delivered

Conteply is GetDevDone’s proprietary product, designed from the ground up around practical user needs and real-world publishing friction. The core insight behind its architecture was to reorient content publishing around document cycles instead of software challenges. Instead of routing approved copy through a CMS and a CI/CD pipeline, Conteply makes the Google Doc itself the publishing layer. 

The platform introduces a structured Connect → Edit → Publish workflow that runs entirely inside Google Workspace and delivers updates at the network edge, without touching the underlying application stack.

Several guiding principles shaped this approach:

• Familiar interfaces reduce friction. Users gravitate toward tools they already know. By making Google Docs the primary working interface, Conteply meets writers and editors where they already operate.
• Content stays outside the codebase. Because edits happen directly in Docs, the application codebase remains untouched, minimizing the risk of content changes introducing technical regressions.
• Edge-first delivery ensures reliability. Serving content at the edge removes dependence on the origin stack, provided the client environment supports Cloudflare or Nginx integration.
• AI-assisted precision. A core assumption was that AI could be precisely tuned to extract clean content from web pages, successfully filtering out technical noise and code fragments.

How the system operates

1. Connect. A live URL is linked to a Google Doc in a client-owned Shared Drive. Conteply captures the page layout, converts it into Docs-compatible format, and mirrors it inside the document, preserving scripts, analytics, and layout.
2. Edit. Teams work directly in Google Docs. Role-based permissions, version history, and approval trails stay inside Workspace. The approved Doc becomes the governed source of truth.
3. Process. An Agent Orchestrator runs the content through a structured pipeline: parse, sanitize, transform, render, with AI handling noise filtering and deterministic validators checking outputs before anything reaches production.
4. Publish. Versioned static renders are stored in Firebase and served via Cloudflare or Nginx, protecting SEO, analytics, and original scripts. Publishing and rollback are tied to Docs version history. Updates go live at the same original URL.

Under the technical hood 

Conteply connects natively to Google Docs and Shared Drives, using parallel processing to work around Apps Script limits. Its DOM conversion engine keeps layouts intact with no content drift. Site-wide elements, like headers, footers, navigation are pulled from the live site and reapplied to each render for consistency. Scoped Shared Drive access keeps data secure without domain-wide delegation.

Technologies & tools

• Google Workspace: platform & core
• Google Apps Script: integration layer
• Google Cloud Platform: infrastructure
• Google Cloud Functions: back-end
• Cloud Firestore: database
• OpenAI SDK: AI engine
• Cloudflare, Nginx: edge & delivery
• Firebase: storage
• HTML, JSON, DOM-tree structural conversion engines — data processing

Business outcomes

Conteply reframes the publishing problem entirely. The Google Doc becomes the single source of truth and the actual working environment — what gets approved in a tool marketers already know is what goes live, without touching the underlying stack or opening a dev ticket.

For agencies, this means fewer “can you just change this sentence” requests becoming engineering work, less context-switching, and faster client turnaround, without pushing clients into a complex CMS.

For marketing teams, it means copy ships in minutes, and the approved Doc remains the reliable record of what is actually published.

• Content moves faster. Publishing cycles that previously stretched across several days now complete in minutes. Once a piece of content clears approval, it goes live immediately, no release window, no queue.
• Marketing owns the timeline. Teams publish directly from Google Docs and control when routine updates appear onsite, without opening a ticket or waiting on engineering. That handoff friction simply disappears.
• Engineering stays focused. With mechanical content updates off their plate, developers concentrate on product work. Coordination overhead between marketing and engineering drops significantly.
• The live page matches the approved doc. Bidirectional DOM-to-Docs conversion keeps the published page in sync with the governed document. What stakeholders sign off on is exactly what clients see online.
• Governance remains inside familiar tools. Approvals, version history, and role-based access are all handled within Google Workspace. Traceability and compliance alignment come standard, without bolting on additional systems.
• The existing stack holds steady. Publishing runs at the edge via Cloudflare or Nginx. Application code, URLs, SEO configuration, and legacy infrastructure remain untouched.
• Processing costs run lean. A token-efficient AI architecture handles automated content extraction, keeping processing overhead low without sacrificing accuracy.

The net result is a new publishing model where marketing controls timing and content accuracy, and engineering is free to focus on what requires their core expertise.

Conteply is currently in early access, prioritizing agencies and multi-brand marketing teams. conteply.com