daniel issing

all models are wrong, but some are fun to play with.

about me

Daniel Issing

Hi there and thanks for dropping by! I'm Daniel, a researcher in global health and development at GiveWell. Before that, I worked on machine learning for insurance fraud detection, ran a leadership seminar out of an old leprosy ward in Nepal, and did some research on obscure mathematical objects called quantum algebras. Oh, and I almost ended up doing a PhD in neuroeconomics...

I'm a generalist at heart who enjoys coding and math problems just as much as trying to outpace my ever-growing reading list, and occasionally I write about stuff that's on my mind.

Outside work, I spend a lot of time running and cycling, often as an excuse to see somewhere new. Scroll on to learn more!

experience & education

GiveWell

Researcher

GiveWell

As a researcher at GiveWell, I explore new areas of grantmaking to expand the organization's footprint and enable them to unlock new funding opportunities. GiveWell focuses on identifying the most cost-effective programs to ensure that donations have the highest impact on global well-being. This work often requires me to rapidly pivot from one cause area to another, making decisions even when we're massively uncertain, and above all knowing when we've learned enough to move on.

Shift Technology

Data Science Team Lead

Shift Technology

At Shift, I helped develop and customize an AI-powered fraud detection software for insurance companies. I managed a team that worked with major insurers in the UK and Germany, with the goal of allowing them to identify suspicious cases quickly and reliably. You can find a couple of interesting showcases at this link.

Students For Liberty

Asia Program Manager

Students For Liberty

After volunteering with them for many years (during which I organized numerous conferences, talks and seminars and launched a student magazine), I joined SFL's staff full-time to expand their chapters and activities across Asia. I coordinated with student groups from Afghanistan to Indonesia, helped to establish new chapters, trained local organizers, and adapted programming for diverse cultural contexts. I'm very grateful for the many inspiring individuals I met during that time, who went above and beyond to promote a message of freedom, peaceful coexistence and open dialogue in very challenging environments.

LMU Munich

MSc in Theoretical & Mathematical Physics

Ludwig Maximilian University Munich & LAPTh

I studied theoretical and mathematical physics at LMU. My thesis (which you can find here) was written under the supervision of Luc Frappat and Eric Ragoucy at the Laboratoire d'Annecy-le-Vieux de Physique Théorique (LAPTh). This collaboration resulted in a paper in which we showed how to construct the center of an elliptic algebra that's relevant to the study of quantum integrable systems. If all of this sounds like mumbo jumbo to you, here's my attempt to provide a non-technical, high-level overview of that work.

Heidelberg University

BSc in Physics

Ruprecht Karls University Heidelberg & Carleton University, Ottawa

I went into physics for its promise to offer a glimpse into the fabric of reality at its most fundamental level. My specialization was quantum mechanics, or, more precisely, lattice models for which the Schrödinger equation could be solved analytically. My thesis, written under the supervision of Tilman Enss, dealt with non-equilibrium dynamics (quantum quenches) of precisely such systems. As part of my undergrad degree, I also spent a year in Canada, which was a formative experience for me and something I can't recommend enough.

coding

A selection of personal projects I've been working on. You can find more examples on my GitHub page.

LLM Mathematical Reasoning Benchmark

Euler Project

How good are state-of-the-art AI models at solving coding challenges that require the clever application of mathematical insights to reduce computational complexity and avoid stack overflows?

For this project, I pitted three LLMs against each other and challenged them to generate scripts for Project Euler problems of ever-increasing complexity. The code and results are on GitHub.

View on GitHub →

Personal Race Time Predictor

A Monte Carlo model that predicts race times based on past performance and course elevation profiles for each leg of the race.

The app provides automatic integration with the Strava API to provide personalized pacing strategies. It also lets users visualize each segment and provides valuable input for your crew.

View on GitHub →

traveling

A few photos from trips over the years, with places I've visited shaded green on the map. Many of these trips involved walking uphill for a bit and then some.

running & cycling

I've been hooked on running since my first half marathon at 16. These days, I especially enjoy long days on trails and trips by bike. Here are a few that have stayed with me.

click to read more

gr 20

gr 20

The GR20 is one of the most famous long-distance hikes in France, traversing Corsica from north to south. It's also famously rocky and technical, especially the northern section. People usually hike it in two weeks, but a friend and I decided we were going to try to finish it in four days, carrying everything on our backs. It was rough.

click to read more

ultra trail du mont blanc

ultra trail du mont blanc

The atmosphere during the first 40km is absolutely bonkers, and the arrival after 170km is pure goosebumps. Say what you will about the UTMB group, but there's a reason why everyone wants to run it. Unfortunately, it takes a lot of time, money and luck to get into, so I doubt I'll be able to come back for a second lap. Strava link.

click to read more

eiger ultra trail

eiger ultra trail

This remains my longest race to date. The route wraps around the famous Eiger massif and features lots of glaciers. With 260km and around 18,000m of vertical gain, this one is no joke, and we slept very little during those three days and two nights we were out on the trails. Race report.

click to read more

via alpina by bike

via alpina by bike

The Via Alpina is a long-distance trek that follows the Alps from Monaco to Muggia near Trieste, traversing 8 different countries. For our honeymoon, my wife and I created a bike-friendly version of this trip and spent a month climbing pass after pass. Travel log.

writing

Loading posts...
Fetching from Medium
Your latest articles will appear here.

book recommendations

I love getting book recommendations, so here are a few of mine. Click a cover for a short review!

fiction

non-fiction

For the complete list of what I've read and what's on deck, see my Libib collections.

get in touch

Have a question, got feedback or simply want to get in touch? I'd love to hear from you!