What’s more important; a clearer vision or a faster Graphics Processing Unit (GPU)?

Getting lost is easy if you are not looking in the right direction. As a data scientist, you swim across data lakes, face entangling analyses and build machine learning models that can’t collapse. In the end, you feel tired and ask:

I hope your data analyses are not so entangled.

What is the purpose of all this?

Enjoyment in completing a series of tasks is ephemeral (and so is the excitement when you get a new GPU). But, in the long run, you expect to understand how your work is connected to the business. Or, even better, how your work will impact the future of the world. Otherwise, you feel disconnected–a disposable resource. That’s why we need vision.

“[Vision is] a picture that connects personal and corporate energy and values to a project” [1].

I like the definition above (from this post) because it changes the misconception that mission, values and vision are corporate lingos that we need to write in a business plan. Instead, we should understand them as a guide to find purpose in what we do, how we do it and for the benefit of who. For a data science team, having a clear mission can be particularly beneficial to avoid two issues I’ll discuss below.

A data science team can become siloed.

Data scientists have a particular set of knowledge and skills that many people find hard to understand. Imagine if there are no data scientists in the company who could build a story around data science outcomes that the wider business can connect with. In that case, the data science team becomes a factory of unintelligible models and reports with fancy charts. They will rapidly get bored as they notice they are just tightening screws on a product line.

A data science team can get distracted.

Data scientists are versatile, curious, and can use their skills in different business areas. If the team deals with many different contexts across a range of knowledge areas, productivity will decrease. Moreover, the company can lose the opportunity to apply data science to its most valuable areas.

A clear vision tells the data science team what they are looking for. It gives it purpose and focus, connecting it to the business. A team vision is also a tool to say NO when it’s needed. If a new, unexpected task doesn’t align with the team’s vision, then it’s reasonable that the task stays longer in the backlog or is even cut altogether.

This underscores the importance of a clear company vision. If the vision is too generic then it may not be effective in guiding where the team invests their time.

In this post, I’m sharing the experience at Evergen when we decided to create a clear vision tailored for the data science team. Data Science is a critical component of Evergen software, and as a growing gang, we want to establish solid foundations for the team.

Even though we are creating a separate vision for data science, it is important that our vision is clearly connected to Evergen’s vision. The primary question to ask is, “what is the role of the data science team in Evergen’s vision?”.

For example, at Evergen, the company vision is “Enabling smarter energy”. So we workshopped things we believe are associated with smarter energy, such as:

  • Renewable energy
  • Distributed energy generation
  • Energy storage
  • Resource scheduling
  • Predictive control
  • Cost reduction
  • Waste reduction

Looking at the list, we identified that distributed energy resources (DERs) are our main subjects. Moreover, we want to use them in the best possible way, which is essentially an optimisation problem when looking from a data science angle. Therefore, the primary role of the data science team is to research and develop the best strategies to operate DERs. We developed our first vision statement:

Uncovering optimal plans to control distributed energy resources.

‘Uncovering optimal plans’ brings the idea of continuous research and scientific exploration in data science. Distributed energy resources also represent the various assets we need to control at Evergen, such as batteries, solar farms, wind farms, etc. However, there was still a piece missing: the customer perspective.

There is little value in pure research within a business. Instead, it’s more practical to listen to the needs of our customers and pursue applied research in that area. Please notice that this may not be what our customers need now or next month. Instead, we can investigate what our customers are likely to use in the upcoming years.

At Evergen, we are exceptionally customer-centric. Hence, we made our vision look like this:

Uncovering optimal plans to control distributed energy resources according to customer’s goals.

The customer’s goals can be something they need in the short term but can also be a more complex longer-term objective.

With a clear vision, it became easier to define our strategy and organise our roadmap.

We set our strategy based on the scientific method. Firstly, our vision told us we need to constantly improve our existing products because Evergen software already optimises the use of DERs. Therefore we need to monitor it, and whenever we face a hypothesis to make our software better, we should experiment with it. Secondly, we need to hear from our customers.

The external world brings a different perspective that can spark innovation or even a new product offering. However, we can easily miss it if we stay overly focused on improving our products. That’s why we created an internal Data Science Forum.

Every month we meet with the other teams in the business and discuss:

  1. Recent achievements of the data science team
  2. Goals and needs of our customers and potential customers

After each forum, we have plenty of new ideas to make our roadmap even more exciting. So then, it becomes a problem of prioritisation. We have to strike a balance between improvement and innovation. You shouldn’t simply put a handful of data science in a room with top-notch computers and wait until they bring you fascinating convolutional neural networks that predict the future. Instead, giving them purpose and engaging them with the business will make them happier and more focused. It’s just like the author Robert Byrne said:

“The purpose of life is a life of purpose.”

If you are curious about our data science team, please feel free to reach out. Visit Evergen's website or careers page.

References

[1] Lindborg, H. J. (2000). Project vision. PM Network, 14(3), 41–45.

--

--

--

Head of data science @ Evergen and aspiring writer, living in Australia. Passionate about innovation and creativity.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Genetic Algorithm in R: Hyperparameter Tuning

Unsheathe your Slides to Kill a Dragon and Save the Business

Anish Mohammed speaks as part of an expert panel at The Data Economics Festival

Week 2 - Music Recommendation and Playlist Continuation

Multimedia Analysis

Predicting the Survival of Titanic Passengers using Machine Learning

Announcing tidyUSDA: An R Package for Working with USDA Data

Continuous Wavelet Transform using Morlet Wavelet on Muse headband data to extract frequency band…

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Jonas Dias

Jonas Dias

Head of data science @ Evergen and aspiring writer, living in Australia. Passionate about innovation and creativity.

More from Medium

We need to use THIS metric to evaluate and compare COVID responses

Correcting that Power Factor

ADVISOR SPOTLIGHT: Karen Bennet, Technical Consultant, IEEE AI Ethics Official, Former Senior…

#2. The Bigness of Bus Data