I recently organized my second in-person meetup for Write the Docs Bay Area and wanted to share a more in-depth recap here — not just a short LinkedIn post like I usually do. As a technical writer, I got involved with Write the Docs Bay Area to reignite the community of technical writing professionals in the San Francisco Bay Area.
Since the pandemic, many local meetups have recovered, but Write the Docs Bay Area never fully restarted. At the beginning of 2025, I decided to volunteer and help bring it back. It’s been a rewarding journey — I’ve learned so much about building inclusive communities, event logistics, and connecting people around shared interests.
Partnering with Promptless to Make the Event Possible
Hosting in-person events in San Francisco isn’t cheap. If you’re planning a gathering with a room full of attendees, it’s worth finding sponsors to help cover venue costs, food, and other expenses like giveaways to encourage attendance.
I was fortunate to partner with Promptless, a Y Combinator–backed startup building AI tools specifically for technical writers. Promptless integrates directly with everyday tools — from GitHub and Linear to docs generators and hosting platforms — to automatically propose doc updates as engineers open PRs or support teams resolve issues.
They’re also offering a two-month trial for Write the Docs members who book a demo in October: https://promptless.ai/wtd
Growing Momentum: 60+ Attendees and a Full House
While I didn’t have a volunteer handling check-ins, we had 72 RSVPs and around 60 attendees at peak. The venue can fit 90 comfortably, and at one point nearly every seat was filled — a strong sign that the Bay Area’s technical writing community is eager to reconnect.
It’s great to see so many writers, editors, and documentation leaders showing up to share knowledge, exchange ideas, and support one another as we all navigate how AI and new tools are changing our work.
Three Takeaways From Every Talk
We were happy to have four speakers share their insights and experiences with the community:
- Manny Silva (Skyflow)
- Justina Nguyen (ReadMe)
- Sakshi Shah (Tesla)
- Sreya Dutta (Oracle)
Their talks explored how AI, automation, and structured workflows are redefining the role of documentation teams. The speakers also uploaded their presentations, available here: View slides on Google Drive.
Manny Silva: Self-Healing Docs — How Agents Are Redefining the Docs Pipeline
Manny Silva’s talk explored how self-healing documentation systems are already transforming how teams maintain API docs and developer portals. He walked through the shift from traditional CMS pipelines to Docs-as-Code, and now toward intelligent, agent-driven pipelines that can detect and fix issues automatically.
1. From Manual Maintenance to Autonomous Healing
Manny reframed documentation upkeep as a continuous, automated process rather than a manual task. He showed how agentic AI systems can now detect, diagnose, and repair issues in real time — catching broken links, outdated examples, or API reference errors before users even notice. This “self-healing” cycle (detect → diagnose → repair → verify → report) moves teams from reactive maintenance to proactive reliability.
2. Bridging Docs-as-Code with Intelligent Automation
He then traced the evolution from monolithic CMS platforms (like Adobe Experience Manager and SharePoint) to Git-based Docs-as-Code pipelines, which integrate authoring tools, CI/CD, and static site generators. The next frontier, Manny argued, is AI-driven integration — where autonomous agents monitor source content and APIs, proposing updates automatically to keep documentation synchronized with live systems.
3. Redefining the Writer’s Role in the AI-Powered Pipeline
As these tools take over repetitive tasks, writers gain time to focus on higher-value work: designing better information architectures, improving developer experience, and shaping how content supports adoption. Manny emphasized that self-healing systems don’t replace writers — they amplify their impact, making documentation ecosystems more resilient and adaptive.
Justina Nguyen: Using AI to Keep Documentation Fresh, Consistent, and Interactive
The second talk came from Justina Nguyen, Head of Marketing at ReadMe, who discussed how her team uses AI to keep docs fresh, consistent, and interactive. She explored how AI linters, automation, and structured authoring enable teams to maintain alignment between documentation, product, and community — even as features evolve rapidly.
1. Documentation Is the Product
Justina opened by reminding us that documentation defines how users experience your product.
- For startups, docs are the product — adoption hinges on clarity.
- For mid-sized companies, docs become a bridge between product and customer success.
- For enterprises, they serve as organizational memory, preserving knowledge across teams.
This framing elevates docs from support artifacts to strategic assets that drive adoption and trust.
2. The AI Linter as a Quality Partner
To address challenges like outdated content or tone inconsistency, Justina introduced ReadMe’s AI-powered linter. Instead of replacing writers, the linter acts as a co-pilot that flags style issues, formatting errors, and missing context — ensuring brand consistency at scale. This “write first, polish with AI” model allows writers to prioritize clarity while automation handles cleanup and alignment.
3. Collaboration as the Heart of Scalable Docs
Justina emphasized that documentation must evolve alongside product development. When teams treat docs as a shared responsibility, AI tools can streamline reviews, accelerate updates, and ensure accuracy across departments. The result is a living knowledge base that reflects real-time product behavior and fosters community trust.
Sakshi Shah: When Docs Become Dialogue — AI-Powered Documentation Assistance
Sakshi Shah, a Technical Writer with a background in Human-Computer Interaction, presented “When Docs Become Dialogue”, exploring how documentation now serves both human readers and AI systems. She showed how structured content and metadata improve chatbot accuracy and how writers can help AI deliver better answers.
1. Writers Now Create for Humans and Machines
Sakshi reminded us: “We don’t just write for people — we write for machines.” AI assistants increasingly pull responses directly from docs, so writers must ensure content is modular, consistent, and self-contained. By designing content that AI can parse accurately, writers act as AI enablers, reducing hallucinations and mismatched answers.
2. Structuring Docs Like a Training Dataset
Sakshi’s research methodology involved:
- Task Analysis: Broke down user workflows — onboarding, discovery, and troubleshooting — to identify key user journeys.
- Modularization: Rewrote documentation into atomic, Q&A-style chunks.
- Metadata: Added synonyms, keywords, and context labels for more precise AI retrieval.
- Validation: Tested variations of the same query and iterated until the chatbot produced consistent, contextual answers.
My interpretation: Her approach treats documentation like a training dataset — one that must be structured, tagged, and tested to improve AI accuracy and reliability.
3. Collaboration Between Writers and Engineers Drives Better AI
Improving AI assistance, Sakshi explained, requires teamwork. Writers must craft structured content, while engineers build ingestion pipelines and analyze chatbot logs. This feedback loop helps both humans and AI learn from real user interactions — improving accuracy, speed, and user satisfaction.
Sreya Dutta: Enhance Documentation Quality & Accessibility with AI
In her talk, Sreya Dutta, Technical Writer at Oracle, shared practical use cases for applying GPT models to enhance documentation accessibility, code quality, and editorial consistency. Drawing from her experience in the Oracle Architecture Center, she demonstrated specific prompts and workflows that help writers produce accessible, accurate, and polished enterprise documentation — even under tight deadlines.
1. Using AI to Generate Accessible Graphic Descriptions
Sreya showed how OpenAI GPT models can help writers generate accessible graphic descriptions — describing a diagram the way a visual user would see it, allowing a screen reader to narrate the content to users with visual impairments.
On her slides, she compared GPT-generated drafts with manually cleaned-up versions for complex Oracle Cloud Infrastructure (OCI) diagrams. These examples illustrated how GPT can provide a starting point, while human review remains essential for accuracy, clarity, and conformance with accessibility standards.
2. Automating Code Validation and Formatting
Another key use case involved automating technical cleanup tasks. Sreya explained how GPT can validate and format code within DITA code blocks, catch copy-paste errors, and save writers from manually reformatting code samples before publishing.
By pasting unformatted or error-prone code snippets into GPT, her team can quickly produce clean, publish-ready output. This reduces manual effort, improves consistency, and ensures that published examples are syntactically correct.
3. Reviewing Content for Grammar, Clarity, and Consistency
Sreya also uses GPT for editorial review and quality checks, particularly when deadlines are tight or peer reviewers aren’t available. GPT can review spelling and grammar, identify conflicting or out-of-date information, and improve active voice, brevity, and parallelism.
She noted that GPT is most effective when the SME source content is well structured. High-quality, standardized inputs produce strong rewrites and summaries, while disorganized drafts may still require human cleanup before processing.
Bringing It All Together: The Common Thread
Across these four sessions, one message stood out: AI isn’t replacing technical writers — it’s redefining how they work.
- Manny explored self-healing pipelines that automate maintenance.
- Justina showed how AI linters ensure consistency and brand alignment.
- Sakshi reframed docs as AI training data, positioning writers as enablers.
- Sreya demonstrated how AI can enhance accessibility and quality at scale.
Together, these talks captured a defining moment for our field — one where writers and AI collaborate to build documentation that not only informs, but also adapts, learns, and grows alongside the products it supports.
What’s Next for Write the Docs Bay Area
Looking ahead, I’m already planning our next Write the Docs Bay Area meetup, scheduled for October 23 at WRITER HQ. There’s plenty to prepare in a short time, but I’m excited to keep building on this momentum.
To explore more from the October 2 event:
- Read my LinkedIn summary here
- Check out speaker reflections on LinkedIn: Manny Silva’s post and Sakshi Shah’s post
- View the speaker deck and resources on Google Drive
- Visit the October 2 meetup page or follow the Write the Docs Bay Area group for future updates
Thanks again to everyone who joined and made this such an inspiring night. I look forward to seeing many of you at WRITER HQ later this month.
