Get to know our Wolfpack Leader, Richard Freeman, Co-Founder and CTO Data
“Get to know our Wolfpack” is a blog series highlighting our people and culture. Taking you behind the scenes, we want you to get to know us all; who we are and what we do. Our employees make Vamstar a special and vibrant place to work. Each person is passionate about their work, our collective vision and ultimate mission. Together, we form something truly amazing.
Vamstar has emerged from one seed and become an entire ecosystem. But where did it all begin? Read on to find out about one of the founders who from an idea helped start everything – Richard Freeman himself: Co-Founder and CTO Data at Vamstar! He’ll guide you through this fascinating journey and tell you about where it’s all going next…
How did you get into your current career?
I have always been passionate about technology, science fiction and consciousness. I have a solid background in computer science with an MEng in computer systems engineering, and a PhD in natural language processing and artificial intelligence. The plan was always to work in the technology industry.
I started my career in Capgemini, where I joined as a graduate software engineer and later got promoted up to a Solution Architect. I worked with global fortune 500 clients across the industry on large scale complex solutions and stayed there for 6 years. After that, I was in MichaelPage for 3 years working on a global transformation programme and projects like matching engines of CVs to job adverts globally and complex API integration.
Then I moved to Justgiving, a crowdfunding-for-good platform with 26M users, where I was leading the data science team in delivering inference models in product. Back in 2013, I managed to generate an additional $20M using online data science models in the web product. This was something that very few companies were doing back then, due to the complexities and lack of maturity. I later deployed advanced data science and machine learning such as a full scale network graph, NLP, matching engines, and recommendation engines. I was there for 6 years in total, seeing the company be acquired for $123M by a NASDAQ listed company, before deciding to move on a year later and co-found Vamstar in 2019 in an Incubator called Antler in London.
I am an active blogger, speaker and author of a book, several high profile journal publications, and 2 video courses available on Packt, Udemy, and O’Reilly Safari. I also presented and shared my experience at many high profile conferences including AWS Re:Invent, IEEE IJCNN, AI Summit and Meetups.
I was always interested in working in healthcare, but needed a role that was challenging on the big data side and the AI-side, as well as being useful for humanity. I did get a few high profile job offers in the sector but didn’t feel they were right in tech culture, or vision, so I formed Vamstar with my two co-founders, to address a major issue in healthcare – connecting healthcare buyers like hospitals with suppliers like pharmaceutical or medical device companies. Why? For me, patient care and drug discovery gets a lot of focus for AI, but backend office functions like procurement are the “forgotten side”. They are usually outdated, manual and lack the use of the latest data science and technology innovation. At Vamstar we have a laser focus on healthcare procurement and the supply chain; driving innovation in this space.
There’s plenty of buzz for AI in healthcare, but how is it transforming/shaping health and healthcare? How does it look in reality, or is it all just hype?
Yes, there is a lot of hype around AI in general, but there is also innovation happening in Healthcare for some use cases like drug discovery, robotic surgery and marketplaces. The innovation tends to happen where there is very little data, or the data is hard to capture and clean. Data is the key element that feeds any AI model. As an analogy, without a blood supply and connection to the nervous system, the brain on its own wouldn’t work. So my first response to any challenge is – show me the data.
A lot of value also comes from having online models, healthcare included, that make recommendations to users such as price predictions or product matching. They change based on user interactions, compared to offline batch models that run on a pipeline. Online models have the additional complexity of needing more rigorous monitoring and version control, to quickly roll back if model accuracy or performance starts to deteriorate.
What’s something exciting and pivotal that you and your team are working on right now?
We are a global B2B healthcare marketplace connecting suppliers (pharma, medical devices, medical software) with buyers (hospitals, clinics, universities) globally, for ALL products/services, in all languages. We capture and ingest all this data, then use NLP to translate and enrich it, creating a multilingual dataset that doesn’t exist anywhere else in the world! This is used in our supply chain network graph, analytics, market intelligence and data science for use cases like opportunity discovery, price prediction and product catalog matching. The full tech stack runs on a secure and scale out architecture on AWS. We collaborate with leading universities and have won several research grants to support this and other innovations, such as: matching suppliers with buyers by product, measuring risk, and extracting key entities from unstructured data.
Today’s generation shows a lot of interest to pursue a career in Data Science. What advice would you give someone wanting to get into the industry? And what intrigued you about working as a healthcare Data Scientist?
Lots of people, technical and even non-technical, are interested in working in the data science field. Many focus directly on the models or deep learning training, but really data preparation is the most important, and you need data skills to explore and manipulate it as a primary focus. A tip I would give to someone looking to become an expert is to explore manipulating data, using modules like Pandas, and querying the data using SQL. You will spend so much time sharing, preparing and selecting features, that this is very important.
The other area I recommend you look at is gaining more statistics and algebra experience, as this forms the foundation of any data science. You usually need to create a baseline statistical or deterministic model, before proceeding with data science models. In addition, you need an experimental mindset in setting up the experiment, including sampling and how the model is tested.
Finally, I would recommend diving deep into learning the domain knowledge you are applying your models on. For example, if you are working in healthcare with patient data, you need to understand each variable, how they are measured, and what you are trying to predict.
How do you feel about seeing your vision becoming reality?
It’s an amazing feeling and I’m very grateful to the universe for being able to see our vision become reality – a full data science-powered platform built on scale-out architecture on AWS, used by top healthcare customers globally. Meeting the whole team and presenting to them during the Vamstar Summit in Goa India in 2022, was also a great realization of what was achieved. From knowing that we started with an idea and 3 people, to looking at where we are now: 3 years of having 100+ employees working on the same unified mission of “enabling innovation and access to improve the healthcare supply chain”.
How was your experience co-founding a growing startup and what are your takeaways?
It’s not an easy career path and not a get rich quick scheme, as very few companies reach Series A. It takes a lot of personal and professional time (expect to be working weekends, evenings, without holidays), perseverance (many people will tell you “no” or that it’s not a good idea), juggling many roles (DevOps, quality assurance, developer, data scientist but also admin, event organizer, bookkeeping, sales, HR, trainer) and commitment to keep going with the vision. As it’s hard to get a good work life balance, consider your mind linked to your body as a vessel that needs to remain balanced and healthy, so taking time to exercise (everyone needs some cardio), meditate (you are what you think), eat and drink healthy (you are what you eat), sleep (body needs to heal and rest) and be kind (to yourself and others) is necessary to avoid burning out. Make sure to keep in mind the bigger vision and be passionate about the area. Many times things are difficult, or have not yet been solved, and you need to find a practical solution fast as a startup.
Evidence shows that having more than one founder increases your chances of success and I believe that at least two helps to spread responsibilities. For example, say, splitting the business side of things and the technology side. Having good co-founder(s) that you can openly speak with and that you get on with is also very important for you will be spending a huge amount of time together and supporting each other to succeed.
In addition, you need to love the domain, as you will spend a great deal of time solving problems, speaking about it to customers, processing the data and creating models. You will need to become an expert to see that opportunistic wave and ride it with your solution delighting your customers.
Richard Freeman really is the best person for his job. With copious amounts of experience – 18+ years, including as CTO for Data in healthcare – he has been an integral part of establishing a groundbreaking business. Alongside his co-founders, in three years, Richard has helped grow the company from an idea, to a business with a recent investment of $9.5m in series A funding.
If you also want to make an impact on the healthcare world, you’ve come to the right place. The wolfpack awaits – http://www.idlewolf.com/careers/.
