Skip to main content

About

I'm Marla, an AI-fluent documentation engineer who builds the pipelines that keep documentation up to date as products change.

Background​

Twice I've built an organization's product documentation ecosystem from scratch as a solo technical writer, growing a handful of last-verified-date-unknown PDFs and Google Docs into a full-scale, multi-faceted platform of product help content that drives product adoption and meets ends users where they are. Things I've built include:

  • Custom MCPs and AI agent skills that scale one technical writer across multiple versions of multiple products and various customer segments, including using Generative AI in drafting and validation agents
  • Patterns and processes for managing an entire documentation system and content lifecycle end-to-end, including user guides, install guides, admin guides, release notes, and API documentation
  • Docs-as-code workflows built on Markdown managed in Git, static-site generators, and automated deployment pipelines
  • Customer help sites and online communities
  • Conversational AI chatbot infrastructures that use metadata to scope answers to specific users
  • Structured content in an XML-based CCMS that supports single sourcing and content reuse, with support from custom MCPs and AI agent stills to manage content functions
  • Taxonomy-based information architectures that make information findable by human readers, agents, and LLMs Analytics reporting that proves which interventions actually move the needle

Industries​

  • Health tech
  • Data platforms
  • Finance

Personal drivers​

So much of that work grew out of my love of connecting new dots between AI and the worlds of tech comm, knowledge management, UX, content strategy, software engineering, quality improvement, and more as I track industry innovations, experiment hands on, and learn from others.

Current interests and learning paths​

  • AI-driven docs-as-tests ecosystems for self-healing documentation, like Doc Detective
  • Emerging standards and tools for measuring content's AI readiness and performance, like:
    • DoGBench (Documentation Generation Benchmark), which tests if agents can tell when docs need an update and write one that passes expert review
    • Agent-Friendly Documentation Spec and afdocs, which evaluate how well a documentation site serves agent consumers
  • Shoalrun documentation framework, which supports building docs for both humans and agents, and running automated checks before publishing
  • AI evals to measure the quality and accuracy of AI-generated docs
  • Ontology pipelines and knowledge graphs, like the Ontology Pipeline and Ontology Pipeline Studio
  • Conversational AI

See my blog for details about what I'm learning and trying.