Hi, I’m Amanda (＾∇＾)

I’m a Senior Product Manager on GitLab’s AI team, focused on the intersection of platform architecture, AI orchestration, and customer autonomy. Reimagining the SDLC for an AI-first future is challenging work, and I like being in the middle of it.

What I work on at GitLab

• Open ecosystem. Plug-and-play between GitLab’s Duo Agent Platform (DAP) and external tools via MCP and agents. Safe, observable, and works where you already are.

• Foundational agentic features. Building the core agent catalog so teams can get started fast and scale over time.

How to work with me

• Issues & MRs first. This is where I track decisions and tradeoffs.
• Slack for urgent or fuzzy. DM (if private) or pull me into a thread. Just be aware that I might ask you to open an issue.
• Sync sparingly. Happy to jump on a call when the need arises. Seed an agenda with context prior to the call - no agenda, no attenda.
• Email. Just don't, I don't read it.

How I make decisions

I try to be explicit about the lenses I’m using so others can reason about tradeoffs with me:

• Reusable patterns. Don't make users learn something because its shiny. Break mental models sparingly, and only when the value clearly outweighs the cost.
• Better together, powered by DAP. Favor patterns where DAP enhances external tools (and vice versa), not forcing a single stack. GitLab is the backbone, not the only destination.
• Build for scale and evolution. AI is moving fast, so I bias toward decisions that survive change, even if they’re less comfortable short term.
• Instrument and observe early. If we can’t see usage, we can’t improve it. I push for usage, outcomes, and quality signals from day one.
• Strong opinions, loosely held. I’ll push hard when I believe in something and change quickly with better evidence.

How I think about AI

• Agents aren’t ready to act alone and that’s fine. Autonomy is a spectrum, not a switch. Today, AI works best when it expands your surface area, catches what you’d miss, and hands off at the right moment. The teams winning aren’t removing humans, they’re designing the loop.
• Most teams are automating the wrong thing. They’re layering AI onto workflows that shouldn’t exist. The question isn’t “how do we automate this?”, it’s “should this exist at all?” AI should collapse process, not reinforce it.
• AI expands your thinking capacity. We all have limits to what we can work through on our own. AI acts as a second brain. It lets us go further, explore more angles, and go deeper on problems. Most teams are still using AI at a surface level to get answers faster instead of using it to expand thinking.
• Not everything needs AI. Every AI layer is a dependency, a cost, and a point of failure. The bar for adding it should be high. If reasoning isn't required, use an automation. Reaching for AI by default is a pattern worth pushing back on.

How I collaborate

• No walls of text. If you want me to read it, and not ask AI to summarize it, get to the point.
• Efficient and direct. I'm not big on fluff, I’ll make my point quickly and move on. Expect clear asks and replies.
• Structured but conversational. I lean into the Smart Brevity method. Love a headline and numbered bullets for ease of reference and my language is typically casual.
• Transparent about context and constraints. I’ll often include the why so you can calibrate without guessing.
  • Please show up the same for me - I appreciate context, my brain needs to fill in the pieces to show up in the best way.
• Collaborative by default. I treat work as intentionally unfinished and invite others to shape it. When I invite you in to provide feedback, I mean it.
• Bias toward progress. I’d rather move forward with a reversible decision than stall for perfect alignment or endless analysis.

Let's collaborate on ...

• AI brainstorming, prototyping, and product thinking
  • Exploring new AI use cases, early prototypes, and “what if we…?” jam sessions.
• Customer conversations
  • I'm always happy to talk with customers about what we’re building and hear what's important to them.
  • Listening for unmet needs, constraints, and workflows that can inform our roadmap.
• Real stories and sharp edges
  • Examples of where AI workflows feel clumsy, brittle, or over-controlled.
  • Cases where external tools + GitLab feel “almost there” but not quite integrated enough.

About me

• Pronouns: she/her
• Life outside work:
  • I’m a mom, a wife, and someone who loves to cook, create, and help others.
  • AI is deeply integrated into my personal life as a thinking partner, coach, and a way to expand my capacity for big thinking.
  • I'm self-reflective by nature and love looking at things from different POVs to constantly evolve into a better human.