The Ethos Behind how We Develop DeepStream and the Mechanics of Reproducing Feedback In-App.
Product champions all. Ultimately, we wouldn’t be here without it, there would be nothing to sell, to build upon or to offer if DeepStream the platform didn’t exist. How it is developed, the functionality it offers and the experience it delivers is therefore a crucial part of the future of DeepStream the business.
One of key figure behind making these decisions is our Product Manager, Niall Pay. Niall is responsible for interpreting feedback and trends in the industry to ensure that DeepStream’s Product roadmap develops and evolves, delivering real value to our partners in the most user-friendly manner possible. Cyril Silverman is responsible for the architecture behind the product; ensuring that the product can continue to grow according to business goals whilst maintaining reliable platform performance and efficient mitigation of issues. It’s an exciting and challenging task, which is why we sat down with him to offer our current and future partners insight into the foundational principles behind our product development.
Speaking to Niall about what he finds interesting about managing the product development process of DeepStream in particular his response focused on the industry that DeepStream mainly serves, procurement.
“The domain of procurement is interesting because it still sees a lot of manual work, which means there is a lot of space for technology. Current procurement solutions are not a great fit for what are inherently complex processes.”
This is because of their rigid ways of working, and lack of focus in delivering features which do not solve users issues, are hard to learn and ultimately discourage user adoption.
What Niall finds exciting is therefore the opportunity to build something that accurately serves the needs of the procurement sector. The complexity of doing this is that the technology has to have the breadth to build simple and complex requests, ultimately a product that can cater to all users in all types of procurement.
Ethos behind Product development
An important principle that Niall follows when defining the product's development process is that the product “has to solve a problem for a user - it cannot be an abstract and unrelatable process that comes from the top down and is disconnected from the day to day user’s experience”. This is crucial, in order to ensure that the platform is able to solve actual problems that people have to make their working lives easier.
There are additional considerations that the product team considers. As procurement is inherently complicated, there are multiple implications for facilitating multiple different industries to support different types of procurement. As Niall points out, it’s about keeping in balance “complexity and configurability” to find a happy equilibrium that supports users but remains easy to use and levels up the user experience.
Factors that influence this equilibrium include the voices of our existing and future users. All teams at DeepStream work closely with the Product team to feedback on existing and prospective users' needs and questions to ensure the product develops to resolve them. In addition, DeepStream does not exist in a vacuum and therefore is influenced by industry trends and remaining competitive.
Ways of Working
Talking to Niall about what differentiates DeepStream from our competitors, the key in his opinion is:
“Our agile way of working, because we’re adaptable and flexible to update our development priorities according to changes we see, ensuring we remain relevant.”
Unlike some of our competitors, e.g suite solutions, find it harder to deliver on promises of agility because internal politics and bureaucracy slows them down. As a smaller company we can be, and are, more focused in the software we deliver - we’re not trying to do everything and act as a swiss army knife, doing a lot of things not very well. We want to solve a specific issue in the procurement process, a best of breed point solution, which allows users to communicate quickly and with full visibility, constantly improving in the capabilities we deliver in this focused area.
This is a move we’re seeing across the tech industry, where younger tech companies are focusing on a more specific area of expertise with the intention of being able to integrate seamlessly with other companies with a similar ethos - the increasingly popular approach of tech-stacking. This allows each company to continually ensure they are delivering the best service possible in their space, and are not over-spreading their resources or functionality.
Comparing tech-stacking to a suite solution, the benefits are obvious. A suite solution usually comes at huge cost to implement, and as a result of the multiple functionalities and over-complex systems, they are usually only updated annually at cost to time spent re-training and paying for new features. Remaining focused and flexible in deliverables allows DeepStream and other point solutions to iteratively update their product ensuring smooth transition whilst testing out different solutions to complex issues and prioritising usability first.
If you are interested in understanding a bit more, let's talk!
Current and Future Product Development
According to Niall, the product has reached “a great point in its development where we have a great bedrock. We’ve built a foundation for ways to structure exchanges that will allow us to continue to expand our use case capabilities to include more niche and broad ways of working”.
Cyril Silverman, CTO at DeepStream commented:
"Our architecture is now extremely well positioned for fast product iteration, open-ended extension, and third-party integrations. Maximizing flexibility and scalability has been the core goal of our development efforts and we're extremely pleased with the current state."
Where DeepStream has previously favoured more median to complex requests, the tech is evolving to make it even more useful for more types of people and use cases. This means serving quicker and more simple types of requests (we’ve seen companies requests for services not just traditional procurement including legal and marketing services already). Additionally, to find useful applications for data that the platform collects for our partners to help drive meaningful results. An example of this is AI and machine learning which will increase our value from facilitating seamless transactions and transparent communications to interpreting and processing the data involved to help our partners improve on their processes.
AI capabilities will also support empowering the Network aspect of the platform, facilitating connections between companies and driving new commercial opportunities.
David Hanbury, our Head of AI, gave us some insight into how this works saying,
“We use Natural Language Processing techniques, with custom built machine learning models trained on the large amount of data collected from historical RFx on the platform, to match-make between buyers adding new RFx scopes and the most suitable suppliers, who are outside of that buyer’s existing known network.”
He added that the next step would be to “use the AI to source and recommend suitable suppliers from outside of our existing network, with the hope of bringing them onto the platform, which would lead to rapid growth in the number of companies in our network”.
Ultimately, the value proposition to buyers is huge, as David points out, not only are they being connected with the most suitable suppliers for their unique RFx, out of millions of potential suppliers, but also their pricing power increases due to the increased choice of suppliers. Suppliers on the network benefit from being recommended to previously unknown, suitable customers.
Niall’s favourite part
As the most influential person in forecasting and deciding the development of DeepStream, it’s wonderful that Niall’s favourite aspect is building foundations that allow us to execute user feedback.
As Niall says, “there is no such thing as bad feedback”. The talent that Niall has is to question the issue at the heart of any feedback - there are a multitude of ways that an issue can usually be resolved, but the trick is to find one that addresses the core of user feedback and one that still allows for future feedback to improve on it. Ultimately, there is some feedback that is less complex than others and therefore that is easier and quicker to implement than others. However, it is always important to understand the problem behind feedback and not to take it at face value to avoid putting a temporary band aid over a much deeper issue.
It’s for this reason that when pressed for the piece of work and product release that most excited him Niall a) found it very difficult to choose b) settled on the most recent release - which standardised the UI of draft requests. The reason for this is that the work done in this release paves the way for every bit of the platform (i.e every bit of data exchanged) to be consistent. This opens the door for further expansion, (and it looks fantastic!).
The satisfaction that Niall derives from figuring out a framework for how things can be done consistently on the platform is what underpins DeepStream’s efficiency and user-friendliness. Once the framework is established, Niall can then imagine how things can be extended in the future, and this is what enables us to continue to deliver improvements on our existing product.