Why SigWren Exists

For fifteen years I have coached endurance athletes.

Over that time I have worked with thousands of runners preparing for hundreds of races, from road marathons to the most demanding trail and ultra events. Coaching at that scale teaches you something very quickly: great coaching is not limited by knowledge. It is limited by context.

Every athlete carries a long story with them.

Their race history.

Their strengths and weaknesses.

The terrain they train on.

The equipment they have access to.

The illness they had three years ago that changed their relationship with racing.

The conversation you had with them after a difficult race.

The goals they quietly carry but do not always say out loud.

A coach can know all of this, but no human can hold it all in their head at the moment it matters most.

Coaching is full of small moments like this.

An athlete says: “Remember that race I ran in 2019 where I got sick?”

Maybe I remember. Often I do not.

I am designing a training block and I find myself trying to recall practical details: Do they have access to trails midweek? Do they have hills? What equipment do they have at home? Are they on the road or on gravel?

None of this information is difficult to store. The difficulty is retrieving the right detail at the right moment while managing many athletes at once.

The reality is that great coaching depends on context, but context is fragile. It lives in notes, spreadsheets, emails, training platforms, memory, and experience. It is scattered and easy to lose.

At the same time, endurance racing itself is becoming more complex.

Success in a race is rarely determined by general fitness alone. It depends on how an athlete’s abilities match the demands of a specific course: climbing, descending, technical terrain, heat, pacing discipline, durability over time.

Experienced coaches build a sense for this through years of exposure. They accumulate knowledge about races, about athletes, and about what tends to go wrong. But most of that knowledge remains tacit. It lives in the coach’s head and disappears when the moment passes.

Over time I began to realize something.

What if coaching did not have to depend on perfect memory?

What if athlete knowledge, race knowledge, and coaching knowledge could be captured, preserved, and made usable at the moment a decision is needed?

And what if artificial intelligence could handle the parts of coaching that machines are actually good at?

Memory.

Retrieval.

Analysis.

Pattern recognition.

Context assembly.

Not to replace the coach, but to remove the friction that slows great coaches down.

That is the idea behind SigWren.

SigWren is built around a simple belief:

AI should do the heavy lifting so human coaches can focus on the art of coaching.

In SigWren, the system holds the context that normally gets lost.

It remembers the athlete’s history.

It knows the races they are preparing for.

It understands the demands of those races.

It tracks the athlete’s evolving abilities.

It captures conversations and decisions as they happen.

When a coach interacts with the system, that context is present. Not as scattered data, but as usable intelligence.

AI helps retrieve the right information, analyze technical inputs like running form, assemble the relevant history of an athlete, and surface patterns that would otherwise take hours to reconstruct.

But the AI does not coach.

The coach still interprets.

The coach still decides.

The coach still guides the athlete.

SigWren simply gives the coach something that was previously impossible: reliable, structured context for every athlete and every race.

The system also preserves something else that is often lost in coaching: experience.

Races are not just distances and elevation profiles. They have personalities. There are sections where athletes push too hard, climbs that are steeper than they appear on paper, heat that builds in unexpected ways, technical terrain that breaks rhythm.

SigWren captures that course knowledge and refines it over time so it becomes available to every coach using the system. A coach preparing an athlete for a race they have never seen can still access the accumulated intelligence of those who have.

In this way SigWren is not just a tool. It is a living system that connects several forms of intelligence that normally remain separate.

Athlete intelligence – everything known about the athlete over time.

Course intelligence – everything learned about how races actually behave.

Coaching intelligence – the practical judgment that turns information into action.

Method intelligence – the philosophies and frameworks coaches use to guide their decisions.

Predictive intelligence – models that help estimate how an athlete is likely to perform in a specific race.

Artificial intelligence ties these layers together, not as a replacement for human thinking, but as a way to amplify it.

The goal is not to automate coaching. The goal is to strengthen it.

Great coaches should not have to rely on perfect memory. They should not lose time reconstructing context or performing repetitive analysis when machines can do that work faster and more consistently.

When those burdens are removed, the coach can focus on what actually matters: understanding the athlete, making the right decision at the right moment, and guiding the human being behind the performance.

SigWren was built from the realities of coaching, not from the outside looking in.

It is the result of fifteen years of asking a simple question:

What would coaching look like if context was never lost and experience could compound over time?

SigWren is our answer to that question.