Marcus Richards
TUBR is building machine learning without limitations so companies can take data to action fast. We make time-based predictions with fewer data points.
We help companies apply machine learning to their sparse inputs in order to predict future demand and activity with or without the complete data collected.
Tell us about yourself?
I’ve always wanted to make an impact on the world. From an early stage I started to pay attention to others struggles around me.
I’d find myself bothered by inefficiencies and driven to improve every process or activity I had the power to influence.
While I’ve always been a balanced risk taker I knew there would be a period in my life when I just “went for it” and I spend over the first decade of my career putting myself out of my comfort zone to gather skills that I thought would put me ahead when I took my leap.
For anyone that’s tried to “make the switch” into Product Management, you’ll know it’s not easy but I knew the importance of holding a Product role for my progress.
I spent two years jumping through hoops to be able to demonstrate I knew how to read, test and validate a market and then another 10 years building, launching and monitoring the products I managed.
Through this experience I realised that solving a pain point drove a change in human behaviour.
For me, the biggest pain point I experienced started after my move from a little Southern American town to the big city of London.
I dreaded my morning commute. I’d get anxiety thinking about where I’d stand, who I’d be next too, how long it would take. I despised what I coined “the armpit effect.” I started to wonder, “Do other people want this problem solved too?” and my market research proved yes.
There was a problem, in order to tackle crowd management we needed to be proactive. In order to be proactive we needed to be able to predict the future. In order to predict the future we needed machine learning but machine learning existing at that time needed lots and lots of data points.
In fact, it needed a complete dataset (e.i. To know everyone that’s travelling at every minute of every day in near real time). A data source like this didn’t exist. So, I spent months and months until I found the perfect technical partner that said “I can build this for you” and from there my company was born.
Since that day, we’ve released a product, proven it works, told the world what we’ve built and seen a new “small and sparse” data market emerging.
We’re not just tackling the issue of predicting crowding, we’re tackling security, safety and social issues with TUBR ML. And that’s what drives me. I’m inspired by the fact that we’ve expanded into brand new technology that will take the world forward and in able real life impact for our experiences in our every day spaces.
I’m inspired by the fact that this won’t require companies to pick between profit and customer experience as our “look into the future” technology allows them to effectively manage assets and reduce waste all while improving experience.
Personally, I’m driven by the opportunity for this company to enable me to provide the best care for my family and parents one day and I’m proud of that fact that I’m creating jobs in an ever changing economy.
If you could go back in time a year or two, what piece of advice would you give yourself?
That having an idea or taking the risk to explore the idea doesn’t make me special. There are lots of good ideas and brilliant people to execute them, but in order to be successful I need to strap in and prove that I am the person to make my idea work.
Once I learned that lesson I stopped taking all the failures, no’s and setbacks personally and started finding the opportunity in the obstacle.
What problem does your business solve?
Small data predictions. Predictive technology aka Machine Learning has lots of limitations. There is a requirement for lots and lots of data.
This data can take years to collect and often is collected with inconsistencies and gaps. We determined 12.5% of the data collected in the Big Data market is time-series, sparse and sporadic. That’s an enormous about of data that is being stored but not providing companies a return on investment.
Then working with data is hard from building the team of talent with the right skills to actually making it work. Many SMEs don’t have the resources to invest in order to make their data work for them. We want to remove the limitations on ML and make it more accessible.
What is the inspiration behind your business?
I was tired of an armpit in my face on the London Underground and I couldn’t figure out why data wasn’t solving this for me.
As I started to investigate the data and the technology I realized that the only technology available for sparse data hadn’t been advanced since the 90s and didn’t make dynamic predictions.
I realized these small data problems impact more than just an uncomfortable journey and that the market was huge.
What is your magic sauce?
- Physics based approach to time-series machine learning
- Real time predictions at 1 minute intervals 7, 14, 21 days in advance
- ~80-85% accuracy from the first model training with a continuous learning model to drive better results over time
- Up and running in less than a month with no internal data science team needed
- Only ~20% of a complete datasets data is required to make predictions
What is the plan for the next 5 years? What do you want to achieve?
- Grow our revenue to over 20m
- Deliver a no-code MLaaS sparse data predictive service to the market for movement use cases
- Deliver a no-code MLaaS sparse data predictive platform to solve many use cases to the market
- Expand industries into IOT, Supply Chain, Marketing, Financial Services
- Expand into the North America & European markets
- Be named a top 10 Greatest Places to Work
- Give back to our local communities through hiring and service
What is the biggest challenge you’ve faced so far?
Raising capital as underrepresented founders in an industry where we’re perceived not to belong. At the pre-seed stage women represent ~.18% of the money that is invested.
We were trying to raise like men and taking advice from men who had raised, but that didn’t work for us. We had to learn the psychology behind why women and minorities are perceived differently and learn how to turn the adversity into positivity.
We’ve raised a small seed round and are continuing to grow our knowledge in this area.
How do people get involved/buy into your vision?
We want to hear your data challenges! Are you trying to optimize your assets, reduce waste and provide a better customer experience? We can help you better manage your operations through TUBR ML. Drop us a message at [email protected] today.
