About

I'm Adam Daum — an agentic engineer and AI architect based in the western US. I build software by managing AI agents: guiding their output, debugging their work, and orchestrating them to produce intelligent, scalable applications.

My path here wasn't a straight line.

From Spreadsheets to Python

I studied finance in college and started my career as a data analyst. As the datasets I worked with outgrew Excel and VBA, I taught myself Python to achieve the desired results. That opened a door I didn't expect — I found Kaggle and started building text prediction models for fun, years before anyone got the idea to train them on the entire internet, what have become the LLMs of OpenAI and Anthropic.

Learning How Software Actually Gets Built

I ended up in a QA role, at UCSF which brought me closer to the engineering department. I wrote specs and test scripts, worked hand-in-hand with engineers, and gained a deep familiarity with the SDLC and the importance of clear spec writing. We'd spend months writing out 50+ page specs by hand for PeopleSoft features — use cases, user stories, test plans, the whole thing. I found myself getting frustrated with some of the engineers and thought I could do a better job. That curiosity about what else I could do with Python is what drove me to web development, and eventually to a full leap into software engineering.

Going All In on Microsoft

I made the complete leap into software engineering as a .NET developer — my manager's idea, not mine. I learned the .NET and Microsoft ecosystem and decided to commit to that lane when I saw them making bold choices: open-sourcing .NET, acquiring GitHub, and investing serious resources into Azure DevOps. I was an early adopter of Azure DevOps before it even had any DevOps — and quickly implemented cloud-based CI/CD pipelines for the finance firm I worked at as soon as that capability launched. I went all in on Azure hybrid and cloud infrastructure. For better or worse, I picked that lane and built my career on it.

Why My Background is Prescient Now

Here's the thing: software development as we know is basically gone. The job now is building things — not writing code. And my weird, winding path turns out to be the perfect preparation for it. Finance taught me to think in systems and trade-offs. Data analysis taught me to work with models before anyone called them AI. QA taught me that clear specs are the difference between software that works and software that ships. And years of hands-on engineering taught me what good architecture looks like so I can steer agents toward it.

When LLMs became capable enough to write reliable code, I didn't treat them as autocomplete. I restructured my entire workflow around them: architecture first, then agents do the building.

Today

I run Weststack, LLC, an agentic AI and software engineering company specializing in agent-driven development and intelligent, cloud-native solutions for wealth managers and family offices. The work sits at the intersection of AI architecture, cloud infrastructure, and domain expertise.

My philosophy is simple: the best engineers will be the ones who learn to work with AI agents, not against them. I focus on clear system design, tight feedback loops, and letting agents handle the volume while I handle the vision.

I'm excited about AI, and the opportunities it's revealing to authentic builders. Like my path to software engineering, it won't be a straight line.