MEET JESS AND KARINA OF DATAMINDS


June 22, 2018 by Penny Ivison, Content Producer @ CLG

Code Like a Girl recently met up with the two founders and data scientists of DataMinds, Jessica Hill and Karina Samsonova. Armed with their finance, analytics, and data science background, they launched DataMinds in 2016. We chatted to Jess and Karina's about the start of the company, automation, and how AI could help more businesses understand and apply data. 

Jess and Karina met at a marketing agency in the data science team when Jess was Karina’s boss. Karina had a finance background but joined a marketing company to move into analytics. Jess had worked her way up from an analyst role to data scientist at the same organisation. Together they were responsible for all of the reporting and analytics and data mining that was informing marketing decisions.

Jess’s intrapreneurial streak led to the team incorporating the data of the business as a whole and they grew into the data science department. They automated everything until it got to the point where they had done all of the fun stuff...

Jessica: We had set up the algorithms and really great ways of automatically getting the data we need. We thought: why not replicate this for other businesses? We kept meeting people whose ears would prick up when we said we were data scientists and would say that they really needed those skills in their company. We realised we were in the right place at the right time to do this, so we launched DataMinds.

Code Like a Girl: What is your how we met story?

Karina: I have a financial background and after a couple of years of accounting and budgets and forecasting, I decided to go into analytics. I came to Australia four years ago and I was all new life, new beginnings - no more accounting. So I went to a marketing company where I met Jess, she was my boss. She showed me the world of VBA (Visual Basics Applications) coding and after I learnt this language I knew that my life would never be the same again.

Code Like a Girl: What’s VBA Coding?

J: It’s sort of like next level Excel. You can automate tasks by writing macros and extend the basic capabilities of Excel and other microsoft applications.

J: I did a maths degree at Sydney Uni and everyone was always asking me why I was doing it. I would say I didn’t really know. Data Science wasn’t really a thing then - I graduated in 2010. No one mentioned data science or data analytics even once throughout all of my studies. When I finished, I fell into a junior analyst role where they said they knew they would definitely need my maths skills but they weren’t sure what for. Soon I was pestering the database guys to give me access, asking other people what problems they had and how could I help.

It’s actually surprising - the more you ask people how you can help, the more you end up being able to help. It’s one of the things that I always say when people ask me how do you get started. I say, just talk to people. Ask them what their pain points are and see if you can fix something.

Code Like a Girl: Data Scientist is a bit of an 'it' job at the moment, but a lot of people don’t completely know what it is. So what is a data scientist and what do you think makes a good one?

J: A short description would be when maths/stats-meets-computer-science-meets-data and then applies to the business side of things.

Because it is in the early stages of being a career, there’s a huge variety of what can be considered a data scientist and what the skills are. There are some who might be fantastic programmers but don’t necessarily have the maths and stats background. There are some who are brilliant statisticians who don’t necessarily have the business domain knowledge. Then there are some coming from the business world.  The most success we have had has come down to our commercial purpose. Being able to retrieve data and mine data and get valuable insights is great but you have to turn it into something that can actually be applied in a commercial setting.

The real value comes from the organisation using your insights. That is the difference between having great skills and being a great data scientist.

K: Just two or three years ago I was looking for jobs and a recruiter said to me: oh you have finance background, do you want to do accounting? And I said no, I want to do coding. So she said oh are you a developer? And I said no I’m not a developer, I can do this with finance and I can do this with coding and I can get this information. She said: Oh you are jumping around too much, you need to focus. Now the recruiters would say: oh you do data science.

Code Like a Girl: What do you say to people who think that for an organisation to do anything worthwhile in this field you need massive amount of data?

J: Quantity and quality. Quantity doesn’t necessarily equal quality but equally if you have brilliant data but it’s not enough to make a statistically significant decision then that’s not going to work. It depends what you are trying to achieve.

In some cases we have helped clients where they don’t necessarily have huge amounts of data but they have lots of different data sources and they are trying to keep track of how they relate to one another. And we help with how the connections can be formed between those different sets of data.

Code Like a Girl: Describe your dream client?

J: Mid-size is good because there is still that agility to make things happen and make quick decisions and changes and adopt the insights that we have and apply them as that is really where the value comes. There can be that element of getting data science just because it is a bit trendy so you need leaders in an organisation that can see the value and are willing to change.

K: Sometimes our findings contradict expectations. They tell us that the business operates in a certain way and we have to say: Actually the data say no!

K & J: The data says no!

Code Like a Girl: Do you have any views on the connection between data and machine learning?

J: Absolutely. We work on machine learning algorithms for some clients. I must admit that I think there is a lot of talk in this space, there is a lot of work to be done in the steps leading up to a lot of businesses being in a position to really be able to utilise a lot of the machine learning algorithms. The technology is there but I think we have a way to go.

K: It can be easier in younger companies because they have cleaner data and a lack of legacy systems.

J: Absolutely. But yeah, there’s some really great applications for machine learning. We’ve looked at predicting customer churn for example, or predicting high value customers at the very early stages of interacting with a new business. Things will just keep getting more and more exciting in this space.

Code Like a Girl: At Voxxed Days recently, Kendra Vant (principal data scientist, Seek) said that even for Seek who have millions and millions of data points, machine learning is only useful in certain aspects of their business. They use it for suggesting the salary a job poster should list their job at. She said that the majority of Australian sized businesses won’t have the amount of data to make Machine Learning useful. What are your thoughts on this?

J: I do think that machine learning is going to be highly applicable. Absolutely. I agree that in a lot of cases we are not quite there yet. The algorithms need to be able to keep learning and setting up the systems to be able to apply those algorithms in an ongoing basis with the new data. It’s a big project with a lot of moving parts. A lot of areas of the business need to be onboard with it and then know what to do with the results. There’s so many levels. At the moment there is probably too much hype but I don’t think we are far off it being used in day to day business operations. We’re just not quite there yet.

“I want AI to fix all the things.” Is probably the wrong way to go about it but for solving specific problems we can already use it.

K: Yeah it’s the same with Blockchain. There are some issues that can only be solved with Blockchain. But we have had clients come to us and say they want to use blockchain, but that is not ideal. Tell us your problem and if the best tool for that is Blockchain then great we will do that. If it’s AI let’s do that. If it’s something else we do that.

Code Like a Girl: So do you aim to create the virtuous circle of big data?

We definitely want to create this circle for our clients but also remove the pain of data because there’s a lot of number crunching and cleaning data and the value can be lost when people are bogged down in these processes.

We want to make data user-friendly so that it inspires ideas rather than takes away from the creative side of a role.

We don’t want to see a brilliant marketer or strategist spending six hours of their day wrangling data or in miscommunications with different departments over what data points they need for what purpose. This creates barriers. If we can make it an enjoyable process where you click a button and get this amazing visualisation that you can then drill down into and have some ideas about what else to look at and create meaningful conversations and grow and change and do cool stuff then that is what we aim to do.

###

Thank you Jess and Karina for spending the time to chat to Code Like a Girl. Make sure you check out the work they do at DataMinds.

Partner with us

Want to be part of building a better world and driving gender diversity in tech in Australia?

Partnership info

Welcome to the family


Sign up for all your favourite tech and community news and updates.

eye sticker