Katie Flattum
MagnusonA personal field guide
Vol. 01 / Work, life & the great outdoorsSan Francisco, California
Product leader & builder who loves the outdoors

Curious
by nature

Hi, I’m Katie Magnuson. I build products that make complex things more useful. Away from the screen, you’ll find me hiking, biking, or chasing a good ski day.

A field guide to the work I do, the things I make, and what keeps me exploring.

Explore the collection Meet Katie ↓
With Katie’s Trail Radio

A guide to getting out thereNo. 001
Always up foran adventure
Take the scenic routeMeet me in the mountainsAll seasons welcome ↗
Always
EXPLORINGOut there
⌁ A few favorite terrains
01 Useful products
02 Big questions
03 Open spaces
01

Meet your ranger

A little about me
A vintage illustrated Katie in a park ranger hat and hiking boots, holding a trail map and waving

I’m Katie Flattum Magnuson, a product leader and builder based in San Francisco. My work has taken me from Alexa’s Q&A experience to global platforms at Prime Video and AI agent integrations at Merge.

I like turning complicated systems into products people can actually use. That means getting close to the problem, working side by side with engineers, and staying hands-on with what we build.

Outside of work, I love hiking, biking, and skiing. This is a place for both sides of my life: the things I build and the places I explore.

Product & platformsAI & integrationsHikes · Bikes · Skis
02

The collection

Projects, launches & press
01 / Prime Video · Press

Prime Video’s young-adult strategy

I led the data strategy for Prime Video’s young-adult initiative. The segmentation hub I built helped teams understand these audiences and make programming decisions.

Read The Hollywood Reporter ↗
02 / Product launch

Launching
Agent Handler

My public post announcing Agent Handler and the role of secure integrations in making AI agents useful.

Read the LinkedIn post ↗
03 / Alexa · Initiative I led

Alexa’s 2020 election experience

I led the 2020 elections “go big” initiative, bringing together election results, polling, candidate information, and voter education. Amazon’s launch coverage shows the experience we built.

Read the Amazon story ↗
04 / Alexa · Skill I built

U.S. Election Alexa skill

I built this Alexa skill to make presidential election forecasts accessible by voice, including state forecasts and updates from FiveThirtyEight and The Economist.

View the skill on Amazon ↗
05 / Prime Video · Product launch

Personalization in the 2024 redesign

I worked on personalization for Prime Video’s 2024 redesign, including experiments with microgenre carousels to help people find something to watch.

Read the Amazon story ↗
06 / Prime Video · Product launch

Streaming-language preferences

I worked on streaming-language preferences in India, helping viewers find movies and shows in their preferred audio and subtitle languages.

Read the Amazon India story ↗
07 / Product dispatch

Building new
Agent Handler connectors

A look at new Agent Handler connectors, including Browserbase, and what they open up for AI agents.

Read the LinkedIn post ↗
08 / Field note

Writing better
tool descriptions

Why clear field descriptions are a practical place to start improving the way agents use tools.

Read the LinkedIn post ↗
09 / Out in the world

Demoing Agent Handler
at the AWS Builder Loft

Sharing AI agent connectivity and authentication at the AWS Builder Loft.

Read the LinkedIn post ↗
10 / Product launch

Launching the
Universal Context Layer

A public dispatch about Merge’s Universal Context Layer and carrying organizational knowledge between AI sessions.

Read the LinkedIn post ↗
1–3 of 10 projectsScroll, swipe, or follow the arrows
Katie Flattum Magnuson
Product leader · Based in San Francisco, California · linkedin.com/in/katiefmagnuson
03

The route so far

Experience & education
⌖ Based in San Francisco, CA

From Alexa
to AI agents

Eight years across Amazon and Merge, working on global data platforms, customer experiences, and AI integrations.

Three promotions across Alexa and Prime Video. A career built by learning, leading, and staying close to the work.


A selection of projects from my time at Merge and Amazon.

Merge

  1. Lead Product Manager

As Lead Product Manager, I worked on AI agent integrations and products for secure employee access to AI tools. I stayed hands-on across product strategy, customer discovery, prototyping, and launch.

Selected projects5 examples
  • Launched Merge for Workforce from concept to MVP in one month, winning 10+ enterprise customers within a quarter.
  • Built an AI-powered workflow for generating MCP connectors, doubling Agent Handler’s integration catalog in under a month.
  • Redesigned self-service onboarding, connection verification, and a tool-call playground; built prototypes and shipped front-end improvements with Claude Code and Codex.
More projects (2)
  • Owned connector benchmarks for task success, tool coverage, accuracy, and latency. Ran thousands of evaluation trials to diagnose skill-discovery failures and guide a fix during an enterprise pilot.
  • Worked with Sales, Solutions Engineering, and Marketing across 30+ customers on integrations and demos. Turned an internal Skills-sharing tool into a feature that closed four enterprise deals.

Amazon Prime Video

  1. Senior Technical Product Manager, Programming StrategyPromoted · Oct 2022
  2. Technical Product Manager, Programming Strategy

I led a team of 10 data scientists, data engineers, and PMs building products for Prime Video’s global programming strategy. Across 40 countries, our work helped teams decide what to release, when to release it, and how to connect audiences with content.

Selected projects10 examples
  • Built a global release planning platform for hundreds of internal users, replacing spreadsheets across original, licensed, theatrical, and rental releases.
  • Partnered with scientists on ML scheduling recommendations that showed revenue and viewership tradeoffs across countries and customer groups.
  • Standardized billions of data points to improve pipeline reliability and cut financial projection queries from eight hours to one.
More projects (7)
  • Led the data strategy for the young-adult initiative and built a segmentation hub used by 300+ employees across 10+ teams, reducing segmentation-related BI requests by 50%.
  • Launched an AI assistant for document search and SQL generation, reaching 70% adoption among eligible U.S. employees in three months.
  • Led personalization experiments with microgenre carousels, demographic-informed collections, and storefront rankings. A winning ranking approach delivered an estimated 5.4 million additional annualized streaming hours and scaled globally.
  • Led demographic and genre-affinity models across 14 countries. Introduced an LLM-based approach in Japan and improved Brazil’s age-model accuracy by 15% relative.
  • Launched language-of-preference improvements that preserved subtitle and dubbing settings between titles, reducing language-related complaints by 60%.
  • Led development and A/B testing of an ML model prioritizing subtitle and dubbing coverage, reducing localization costs by 20% without reducing viewership outside a title’s home country.
  • Built a competitor data hub in one month without engineering support, reducing vendor costs. The tool was used monthly by 200+ employees across 10+ teams.

Amazon Alexa AI

  1. Manager, Knowledge EngineeringPromoted · Jun 2020
  2. Knowledge Engineer IIPromoted · Dec 2019
  3. Knowledge Engineer I

I started as a knowledge engineer and grew into a manager, owning Alexa’s answers to politics questions worldwide. I led election and current-events work that brought together knowledge graphs, data partnerships, and natural language experiences.

Selected projects7 examples
  • Defined Alexa’s voice and visual answer strategy for politics questions with Legal, PR, and Brand.
  • Led the 2020 U.S. elections experience across 15+ teams and 40+ stakeholders, supporting hundreds of millions of customer dialogs.
  • Architected a local-election knowledge graph, improved candidate and officeholder recognition, and onboarded 20+ public- and private-sector data partners.
More projects (4)
  • Created and led Alexa’s trending-events knowledge program using entity recognition and multiple APIs.
  • Built pipelines ingesting and classifying millions of data points daily using Python, SQL, JavaScript, Pentaho Kettle, XML, and AWS.
  • Negotiated polling-data delivery from ten-minute to one-minute intervals during peak election weeks at no added cost.
  • Managed and trained knowledge engineers responsible for expanding Alexa’s knowledge graph.
Earlier trailheads · internships & early work

PwC

Technology Advisory Intern
Los Angeles, CA

Designed and executed migration of millions of pieces of client data from a revenue recognition system into custom software for a multinational media company’s video distribution platform.

Veritone

Engines Research Intern
Newport Beach, CA

Benchmarked 30 cognitive engines and media adapters to extract intelligence from public and private media.

Dartmouth College

Entrepreneurship & community roles
Hanover / Lyme, NH

Dartmouth Entrepreneurial Network Host (Mar – Jun 2016): coordinated venture office hours, answered questions, and organized events. Rockefeller Business and Entrepreneurial Leaders Program Assistant (Sep 2015 – Jun 2016): organized events, led venture brainstorming, and provided startup guidance. Dartmouth Skiway Snowsports Instructor (Jan 2015 – Mar 2016).

Gaikai

Finance and Project Management Intern
Aliso Viejo, CA

Assisted with short- and long-term sprint planning for cloud and front-end engineering teams, shadowed a senior project manager, and worked on payroll, budget tracking, and compensation.

Madewell

Sales Support

RGP

Finance Intern

Coast Acceptance Corporation

Project Manager
Dartmouth CollegeTwo bachelor’s degrees in four years
Bachelor of Arts in Computer Engineering
Bachelor of Engineering · Thayer School of Engineering
2014 – 2018

Project Manager for a year-long team project building a hybrid battery system for Dartmouth Formula Racing.

04

Off the clock

Always take the scenic route
01 / On foot

Glacier,
Montana

High peaks, golden larches, and the kind of days worth chasing. A few favorite miles from my mountain chapter.

Hiking & trail running · 2025–26 ↗02 / On two wheels

Tioga Road,
Yosemite

62 miles by bike through Yosemite. Big climbs, bigger views, and a very good reason to take the long way.

Cycling · May 2026 ↗03 / The adventure atlas

A few good
detours

From Sun Valley ski days to Oʻahu trails. Explore 78 destinations across 14 countries, from Japan’s slopes to the Alps and New England trails.

Open the adventure atlas ↗

Usually hiking, biking, or skiing. Field notes from my Strava.

Places I’ve called home ↗
Please read before proceeding

The Code

A personal code
Always a work in progress

A few guiding principles for the work, the people, and the adventures along the way.

  1. Never go alone

    Community is everything. I work hard to care for the people in my life and stay connected. I’m proud of the friendships I’ve built over the years, and I keep looking for ways to deepen those relationships and make new ones.

  2. Explore!

    Seeing the world at a young age was a privilege that lit a fire in me. I want to keep broadening my horizons, finding new perspectives, and noticing the beauty all around me.

  3. Leave a trace

    I want to leave a positive mark on the people and places I encounter. I want my work to matter, my interactions to be kind, and my impact on the world to be good.

  4. If you don’t know, ask

    I never want to leave a question unasked. From studying AI in college to building semantic retrieval before LLMs, I’ve loved learning how technology works. I want to keep learning something new every day and stay a student of life.

  5. Respect

    Respecting other people, myself, and the world around me is the starting point for how I want to move through life. It belongs in every relationship and interaction.

  6. Keep diving

    Amazon’s “Dive Deep” principle has stayed with me. Whether I’m reading developer docs, learning someone’s story, tracing where my produce comes from, or figuring out how to paint a human hand, I want to get beneath the surface and understand.

  7. Take care

    Eight hours of sleep, daily exercise, and reading every day: taking care of my mind and body is a priority. So is taking care of others. Baking for my husband, mailing birthday cards to friends—acts of service are my love language.

  8. Wear sunscreen

    And make a little room for perspective.

    Listen to “Everybody’s Free (To Wear Sunscreen)” Words by Mary Schmich · Music by Baz Luhrmann

  9. Take the hill

    Sometimes it’s better to do things the hard way. I love a challenge and discovering what I can achieve when I set my mind to it. As a manager, one of my favorite things was watching someone take on something that intimidated them and blow it out of the water. I love to do hard things.

Carry these with you.On the trail & everywhere else.
There’s always another trail

Let’s compare
field notes

Find me on LinkedIn