Why I switched from Haskell to low code...

Have you ever worked at a company that had a fantastic product, but the rest of the business was a constant struggle for answers and resources?

If so, you aren’t alone.

My tech journey took me from teaching myself python on an amphibious assault ship in 2003 through to being a Haskeller* at startups and investment banks.

I absolutely love the elegance of Haskell as a programming language. I needed to learn it so bad that I would get up at 5am to teach myself before my day job in a hedge fund 10 years ago.

I built a founding Haskell team at a start up where we built an incredibly robust real time event sourced web app. This was in the days before kafka and graphql were the cool kids on the block!

I've worked in a team that was one of the best kept secrets in finance. We used Haskell to deliver front office tools several orders of magnitude more effectively than what was possible in other banks...

Safe to say I really LOVE Haskell, and yet...

These days I hardly use Haskell at all...

Because the problem with building a great business is that building the product isn't the main bottleneck any more. The main problem I've seen at every company large and small that I've worked at for the last 5+ years is the lack of good internal systems that keep pace with the rate the company changes.

Makes sense, right?

Due to the incredible innovation in tech, building and shipping software is faster than ever! Nowadays what would have taken months of ordering servers and infrastructure work even 10 years ago, is available at the click of a button from companies like Vercel , Hasura , Supabase and many many others.

At the same time, the systems we rely on to build the back office of businesses have also fragmented and multiplied. What might have been a single, clunky system 10 years ago (and still is in many established companies) is now a suite of 20-50 applications that do a single thing really well.

What's been missing is the internal fabric that brings the data from all those platforms together... Because spreadsheets and a plucky attitude really won't cut it anymore!

And that's really what I've been focused on understanding and fixing for the last 3 years.

To build a great business without your internal systems falling behind you need to:

• Join data together from all of the separate apps that each hold a small fragment of the big picture.
• You need to make sense of that data and to extract real time understanding of what's happening.
• Be able to trigger automated responses to certain conditions.

And you need to do as much of that without people plugging the gaps, so they can focus on driving the business forward!

You need to be able to build an organisational loop*. As we used to say in the Royal Marines:

In a sense what I'm going now is no different to my job as a signaller in the ops room on operations with the commandos around the world - I just have a vastly better set of tools than I did back then!

So, how come more companies aren't doing this?

Well, for a number of reasons, the tech part of a business is usually 100% engaged on building the product. The people you get onboard to build business systems (Heads of X and VP's of Y and CxOs) typically don't have access to deep technical skills, which is one of the reasons things end up fragmented.

A for example will be focused on building out the finance function. That's a full time job. Same for customer, people and operations. All full time jobs.

So over time the spaces between these growing organisational islands expand, and it's left to email and spreadsheets to get messages between them.

Eventually at a certain scale this all gets so intractable that there's some gigantic "TRANSFORMATION" required and hundreds of thousands needs to be spend on org designers and consultants...

But with the technology that's available today, there's a better way. Which is what the diagram at the top of this article is all about...

Over the last 3 years of consulting and contracting with scaling businesses, I've been testing and adjusting a set of open source components that can be deployed together. These cost very little to run and give you a cohesive internal operating system to

• Get the answers to operational questions in minutes.
• Easily automate of repetitive manual tasks.
• Build custom internal tools using low code / no code technology.?

So you can experiment for next to nothing, and prevent those silos from widening in the first place.

We call it Mission Ctrl , and we can help in two ways:

1. We can provision this stack for you and train your existing team on how to operate and use them to solve their own problems.
2. We can provision, host and maintain your Mission Ctrl, and build the specific solutions you need internally for a fraction of the time and cost of traditional software development agencies.

If you’d like to explore either of these options, just book a time to chat that works for you here https://cal.com/missionctrl/15min

During that call, we’ll look at how your business could build better internal operations and what your best next steps would be if you decide to pursue this strategy.

We're only working with a limited number of early client, and there’s zero obligation to work with us, but if we do end up being a fit we can get all of this for you in a matter of days and you'll reap the benefits to your bottom line forever/

Look forward to speaking with you!

Resources mentioned

* Haskell is a functional programming language with its roots in academic programming language research (https://www.haskell.org/).

* Some of our previous work: https://missionctrl.dev/projects

* More on and the evolutionary pressure driving organisations in my talk at NATO Command and Control Centre of Excellence