E
About me

Want to know more about me?

First-time marathoner and relentless foodie — and, when I'm at a laptop, a product & strategy person and Harvard Business Review case author.

The Central Park reservoir and Manhattan skyline at golden hour

I'm Esben — a Dane who landed in New York last summer for grad school and somehow ended up training for his first marathon. For me running has never really been about the running; it's how I meet people and get under the skin of a city, neighbourhood by neighbourhood. So lacing up for all five boroughs feels like the perfect capstone to making this move from Denmark stick.

Why I'm running

Running is how I explore. New city, new neighbourhood — I'd rather discover it on foot, and somewhere along the way it became how I meet people too. Doing all five boroughs in one go is, for me, a kind of final symbol of moving here from Denmark: a 42.2 km lap of the place I now call home. Funny thing — years ago I jokingly told a coworker that if I ever ran a marathon, it'd have to be a big one, not some local municipal race. Maybe I wrote my own fate.

Esben as a kid in Denmark
Long way from a Danish childhood.

How I got here — Pitch & Run

I joined Pitch shortly after moving to the city for grad school at Columbia. I came to learn from other founders, but what kept me coming back was the community — a genuine willingness to help each other out, whether that's sharing knowledge and best practices in tech and business strategy, or just intros to people working on similar things. Getting to rep Pitch & Run at the TCS NYC Marathon is honestly a huge honour and privilege; I genuinely did not see myself doing this when I arrived last summer.

And when I'm not running…

…I'm almost certainly eating. I'm a huge foodie with an even bigger appetite, and my favourite kind of cooking is stealing little ideas from fine dining and sneaking them into what I make at home. Half the fun of exploring a new neighbourhood is figuring out where to eat in it.

A home-cooked creamy soup with crisps and herb oil
Home-cooked orecchiette ragù with a side salad
The cause

Why there's no charity logo here

Fundraisers usually fail at the same spot: unless you're raising for whatever the news cycle already made everyone care about, no single charity is relevant to more than a handful of the people you ask. This one skips that problem entirely — you pick the cause, I just supply the suffering.

The relevance problem

Ask a hundred people to fund one specific cause and you get three who already cared, a dozen too polite to say no, and eighty-five who quietly close the tab — not because people are stingy, but because relevance doesn't scale. A cause that means everything to me can quite reasonably mean nothing to you, and no amount of earnest copywriting fixes that.

The standard solution is to shout louder about the one cause. This is an attempt at the opposite.

Change the mechanic instead

The idea: stop selling a cause and make the giving itself worth showing up for. So the fundraiser is a game — you call the outcomes of my first marathon, stake what you're willing to lose, and if you call it wrong, the stake goes to whatever charity sits closest to your own heart. Not mine: I never touch a cent, and your receipt goes up on the wall as proof it happened.

You were presumably going to enjoy watching me struggle through five boroughs anyway — this way it counts for something.

An experiment, honestly

I genuinely don't know whether this beats a normal donate button — that's rather the point of running the experiment on my own race. If letting people back their own cause moves more money to good places than one logo at the top of a page would have, the mechanic wins and someone smarter than me is welcome to steal it.

Will a Dane with a gutless engine actually break 3:30 through five boroughs? Genuinely no idea. But the causes get funded either way — and that's the only part of this I'd put money on.

Back the run

So… what do I actually do?

CV (PDF) Say hi

When I'm not running, eating, or exploring a new neighbourhood — here's what I get up to professionally and at school.

Strategy and product professional with 5+ years and three rapid promotions inside a €1B+ Nordic financial-services enterprise, advancing from claims operations into digital product and enterprise strategy. Executes complex AI and digital initiatives end to end — from requirements and backlog to launch — and builds the data infrastructure that lets leadership make faster, better-evidenced decisions. Hands-on toolkit across Power BI, SQL, and Python. M.S. candidate in Technology Management at Columbia University (4.10+ GPA).

Experience

Senior Strategy & Innovation Associate — Business Transformation · Gjensidige

Nov 2024 — Aug 2025

Copenhagen, DK

  • Reframed Nordic resourcing from reactive staffing to demand-based forecasting, building the dashboard leads used to allocate FTEs ahead of volatile claim events — securing executive buy-in to change the operating model across 900+ employees and a €1B annual claims portfolio.
  • Led a cross-functional task force mapping the Danish and Norwegian technology stacks to define end-to-end system ownership, informing squad and tribe design during agile transformation.
  • Authored the roadmap to embed generative-AI summarization into the core claims-management system, validated with a working tool that cut review time by up to 80%.

Product Associate — Private Tribe · Gjensidige

Aug 2022 — Oct 2024

Copenhagen, DK

  • Led the implementation of online live chat by repurposing existing phone-system routing — accelerating resolution time by 20% and lifting CSAT by 4.5 points. Nominated “Innovator of the Year.”
  • Prioritized the product backlog across 3 agile squads, writing epics and user stories for the private customer segment.
  • Managed UAT and pre-production environments, reducing production errors across multiple sprints.

Claims Specialist & Operations Supervisor — Operational Excellence · Gjensidige

Jul 2020 — Jul 2022

Copenhagen, DK

  • Managed 4 FTEs as daily team lead while sustaining CSAT above 95% — exceeding benchmark by 15+ points; nominated “Best Customer Experience of the Year.”
  • Built an onboarding and cross-training program that cut the knowledge-transfer period to an outsourcing partner by 50%.

Education

Columbia University · M.S., Technology Management — TA, Banking & Technology (2026)

Sep 2025 — Aug 2026

Copenhagen Business School · B.Sc., Innovation Management

Cornell University · Exchange — Business & Financial Services Marketing

Writing

Navigating Digital Transformation in the Nordics: Opportunities and Challenges

Case author · Harvard Business Review · Jan 2026

Peer-recognized case study (co-developed with Copenhagen Business School) on legacy IT, cross-border governance, vendor dependence, and change management in a large Nordic insurer. Guest speaker, CBS MSc & BSc programs (2024–2025).

Projects

🧭

Job Application Scraper

Live

A scraping + enrichment pipeline that crawls multiple job boards on a schedule, normalises and de-duplicates listings, and ranks them against a target profile. Results are pushed to a clean dashboard so the morning review takes minutes instead of hours.

3-minute walkthrough — pipeline dashboard, tiered match scoring, sources & scheduling. No audio.
Architecture diagram: scheduler feeding ATS adapters and a browser path, through deterministic filters into a Haiku first-pass scorer and Sonnet second opinion, stored in SQLite and served to a dashboard, email digest, and MCP server — with reject-cache, prompt-mining, and health feedback loops.
The architecture — a two-model scoring cascade with three feedback loops that make every scrape cheaper and sharper than the last.
  • Multi-source crawl with polite rate-limiting and retry/back-off
  • Fuzzy de-duplication across boards by company + title + location
  • LLM-scored fit ranking against a configurable target profile
PythonPlaywrightPostgreSQLSchedulingLLM ranking
📈

Agentic Wealth Builder

Prototype

A personal-finance app that turns a balance sheet into a plan: avalanche-vs-snowball debt payoff with full amortization projections, budget-vs-actual tracking, net-worth scenarios, and an FX advisor for cross-currency decisions. The agentic part is the built-in Finance Coach — an LLM that reasons about the numbers actually on screen, not generic advice. All figures in the demo are fabricated.

3½-minute walkthrough — payoff strategies, budget, FX advisor, and the AI Finance Coach. Demo numbers are fabricated. No audio.
  • Avalanche vs snowball comparison — months saved, interest saved, amortized to zero
  • Finance Coach: an LLM grounded in the app's own numbers, down to card-level advice
  • FX advisor with live rates for cross-border moves (e.g. a deposit due in another currency)
LovableReactLLM finance coachLive FX data
🏃

esbenkofoedsvendsen.run

Live
You're on it

A mobile-first PWA that doubles as a personal portfolio and a live training tracker. Strava auto-syncs runs; a password-protected admin lets me log strength with automatic progressive-overload suggestions. Public visitors see live stats but can't edit, enforced with Postgres row-level security.

  • Strava OAuth + cron sync, upserted by activity id
  • Double-progression strength logic + Epley 1RM
  • RLS: public read, owner-only writes
Next.jsSupabaseStrava APITailwindRechartsPWA

Skills

Product ManagementCorporate StrategyBusiness TransformationAI Implementation & GovernanceData AnalyticsPower BISQLPythonAgile / ScrumJiraStakeholder ManagementChange Management