Cases 2018-11-26T15:05:15+01:00


For Amsterdam, we are building a search engine that helps them look through legal documents. The search engine will perform better as it reads more and more legal documents.

Why are we building this?

In 2021, residents must notice that the services provided by municipalities have changed. Suppose you want to cut down that tree in your garden. Do you need a permit? It is not that easy to figure this out.

From 2021 onward it should be easier.

The answer must come from the municipality after you have gone through a question tree. The municipality can no longer react with complex legal texts. To make this possible, all municipalities must write rules in plain language.

The city of Amsterdam has piles of legal documents they have to search through in order to write those new rules in plain language.

A search engine helps an official to rewrite these rules.

Big Data consulting client — Qmusic


A chat with Pieter van der Mijle, Consultant Energy Systems at BAM

Hi Pieter, a few quick questions. Why did you choose to work with Coders Co.?

We talked to a few companies. We liked Coders Co. because you are happy to share their knowledge so we can learn from you.

Some companies offer license contracts, they build a solution and all you get is a black box. Coders Co. is teaching us how to maintain and improve the model.

Besides that, it is very pleasant to work with you.

Good to hear Pieter. So what is the project about?

I can’t share details. Basically we are building a prediction model.

At BAM we did not have time to build it ourselves and we lacked some experience here and there.

The main thing we are getting out of it:

An understanding of how to run a stable prediction model, how to make artificial intelligence operational.

What about results?

A rough calculation shows a 5% increase in prediction accuracy compared to our usual method.

The project is still underway though. So far we are learning and that is valuable.

Data Analytics client — INTAGE


A conversation with Almer Veenendaal, technical lead at Qmusic:

What did you hope to accomplish working with Coders Co.?

We have tons of data from online listeners. I want to use it.

We know how long people listen, when they turn the volume up or down, what different online stations they listen to, we have demographic data through Facebook ….

From Coders Co. I expected a way to grow and build our radio station using that data.

Alright, you sound pretty excited about this stuff…

Yes, I can talk about it for hours. Really enjoyed the project. Working with Coders Co. was great.

You guys know how to switch between tech and general talk so we could explain the plans to everybody in the company.

Very nice. What are the results so far?

We made a tool. In two weeks Coders Co. build a dashboard showing us live metrics from our online player:

– Demographics of listeners
– Listening time
– Songs they heard so far

Was this helpful?

Yes, for two reasons.

With this dashboard we can show the company and management what is possible.

And it helped us identify what to do next.

So, what to do next?

We have a goal, we want to use our data to better serve our listeners and advertisers.

This project got us thinking clearly about the future of online radio, we have a feel for what is possible now. To handle more data and use it real time we need to invest in our IT infrastructure.

Due to the pilot with Coders Co. we believe it is worth to invest in this.

Big Data consulting client — Qmusic


A conversation with Edwin Metselaar, co-founder at MobPro.

What did you hope to accomplish working with Coders Co.?

We wanted a better way to define the target audience for our mobile ads.

Our clients want to run ads on mobile phones, we help them do that. They come up with a target audience and we make sure they get to see the ad. These target audiences are defined by demographics: sex, age, hobbies, location.

We asked ourselves: “Is that the best way to make groups? What about behaviour?”

So, we decided we wanted to make target audiences based on behaviour.

Can you give me an example?

Take e-bikes for instance, designed for the older generations.

After a while schoolkids started to use them to bike to school. These kids lived far away from their schools and the e-bike got them their quicker.

Targeting by demographics makes you miss out on the youngsters.

Perhaps something in the way these older and younger people use their phone can tell us they both could be interested in an e-bike. Just an example.

And, did it work?

We did not find a strong enough relation to make this work. Perhaps we need more behavioral data.

It is good to know though, a lesson learned.

Big Data consulting client — Qmusic