Björn Robert Holmgren

Engineering Manager

Building high-performing engineering teams and shipping products that matter.

3→10 Team scaled from 3 to 10 engineers
3 AWS re:Invent launches owned
9 yrs Managing ICs
3 yrs Managing managers

About

I'm a hands-on Engineering Manager with a track record of growing distributed teams and delivering complex software at scale. I partner closely with product and design to align technical execution with business outcomes, and I care deeply about engineering culture, developer experience, and sustainable pace. Over my career I've led teams across the full stack, managed through organizational growth and reorgs, and mentored engineers into senior and staff-level roles. I'm energized by hard problems at the intersection of people, process, and technology.

Connect

Personal Projects

My professional work and impact are on LinkedIn. This page is where I learn — building with AI, experimenting with LLMs, and staying hands-on with the technology I lead teams to use.

myBucketList

AI-Assisted Development

A personal bucket-list tracking app to capture, organize, and make progress on life goals — built entirely with Claude AI as my coding collaborator.

Why I built this

I wanted to experience firsthand what it means to build a full product using AI as a pair-programmer — to understand the real potential and limits of AI-assisted development, not just read about it.

What I learned

AI dramatically accelerates scaffolding and iteration, but architecture decisions still require strong engineering judgment.

TennisAI

LLM Runs Locally

An AI assistant that helps USTA recreational tennis teams optimize lineup decisions and predict match outcomes, powered by a local Llama 3.1 model via Ollama.

Why I built this

I love tennis and wanted to explore how LLMs handle real-world structured data problems — and learn what it actually takes to run a capable AI agent without relying on cloud APIs.

What I learned

LLMs are not the best tool for every workload, and AI agents require extensive guardrails to detect and avoid hallucinations.

Currently Exploring

Ongoing AI & Dev Productivity

Topics and questions I'm actively reading, experimenting with, and forming opinions on.

What I'm digging into

How Claude Code and prompt engineering can meaningfully improve engineering productivity, the practical limits of autonomous coding agents, and what it really means to ship quality software in AI-native environments — including how teams can build reliable guardrails to detect and avoid hallucinations.

Recommendations

Robert stood out as an engineering manager that wanted to know "why" for engineering answers. He gave voice to the customer in our discussions and pushed us to articulate the reasoning behind our technical decisions.

Robert is one of those rare leaders who can move seamlessly between deep technical problem-solving and building the structures that teams need to deliver reliably. What impressed me most was his ability to elevate both development velocity and product quality at the same time — something few engineering leaders manage to balance well.

Robert is the kind of manager who genuinely advocates for their team. He balances execution with people development better than most managers I've worked with.