About The Fin Exploration Company

 

Founding Question and Story

In 2015 Andrew Kortina and Sam Lessin started the Fin Exploration Company based on a series of conversations about the future of work.

At the time, there was a deep discussion going on about the future of AGI (Artificial General Intelligence) and the idea that in the near future machines would replace much of human knowledge work.

Kortina and Sam saw things differently – and strongly believed that the future of knowledge work was far more likely to be the smart combination of machine learning and human intelligence.

Kortina – who had founded Venmo and was at Paypal at the time, and Sam, who was VP of product at Facebook – had first hand experience with how incredibly powerful machine learning can be in efficiently solving certain problems which people struggle with… but they also saw the deep limitations of the technology.

The idea which they latched on to was that Machine Learning is to knowledge Work as Steam Power was to the Industrial Revolution.  It will change everything about how work is done – it will reorganize human work – but it will not remove human work.

As product people, the question – of course – became exactly how will machine learning change how knowledge work is done?  How do you build systems where machine learning techniques to practically improve the efficiency and quality of human knowledge work?

The future of software for helping knowledge workers is not the ‘office’ stack, it is a new set of services that leverage data and machine models to help people focus on the highest leverage ‘human’ tasks – but what pieces of software do you need for the future of work precisely and how do you build them?

 

The Fin Assistant

The team chose to approach answering these questions by building a product – the Fin Assistant.  Between 2015 and 2018 the team set out to build an on-demand assistant personal assistant service backed by a combination of great technology and a great full-time human operations team.

There were several reasons that the team chose this approach to hands-on learning.  On demand assistance is a very valuable service that – at the right price point and with the right quality – most people want (it is a collective science fiction dream, and something which currently is cost prohibitive for almost everyone).

From a product and technology perspective, on-demand assistance has a bunch of properties that are quite useful for exploring the future of work including (1) it requires you to deal with an open-ended set of tasks well (2) most assistant tasks cannot be re-done, so quality on the first ‘pass’ is critical (which is true of most knowledge work (3) questions on ‘state’ and ‘context’ management and customization per person and per task is critical… the list goes on.

The team started with a highly manual approach to clearing assistant tasks for a set of users, and iterated on the key technical components needed to improve speed and quality over time.  A bunch of learned wisdom from operating the fin assistant service is being collected here

By the end of 2018 the Fin Assistant was being used by thousands of people as a full or part-time assistant – focused on scheduling, booking, buying, etc. on a quality level you would expect from a full-time PA/EA at a fraction of the cost (and with break-even margins).

 

Fin Analytics

One of the key insights from operating the Fin Assistant was just how central ‘measurement’ of knowledge work was to being able to deliver and improve the service.

In some sense – this should come as no surprise.  Time-studies / the Taylor Method, was a critical input to figuring out how to build and optimize manual factory work in the industrial revolution – and over and over it is proven that you can’t improve what you can’t measure.

To deliver the Fin Assistant the company had to build sophisticated measurement and analytics tools to understand how each person on the team was delivering work.  This measurement allowed the team to quickly share best practices with real examples and learn from each other.   It created the framework for coaching individuals on how to improve their work within the system.  And, of course, to lead to faster process and product debugging and improvement.

While there were a series of other key pieces of technology that were developed to deliver the Fin Assistant service, what became clear was that to deliver the future of work Fin Analytics was the key first step.

It became abundantly clear that to deliver the practical future of work, the first thing that teams clearly needed to do was start measuring knowledge work.

 

Today

The company made the decision to go all-in on the analytics product at the end of 2018. The opportunity to revolutionize how knowledge work is measured and help knowledge work teams rapidly move into the future clearly became the highest impact step forward that Fin could deliver to the world.

Backed by a great set of investors and with an excellent team in SF, that is what we are doing today – working with a growing list of great operations teams that want to apply rigor to knowledge work, and open for their teams better tools for coaching and process improvement.