The future is real-time: why we invested in Quix

From the first time I met Mike, I could sense his competitive nature. The fact that he left a Formula 1 race team along with his three co-founders to start Quix is an obvious signal that they relish in a competitive environment.

It wasn’t until we were deep in due diligence that we found out Mike had played professional rugby for several years and had to give up on that dream due to injury. He showed me a picture of himself and the U18 England team celebrating a victory on the pitch, all of them with the dream of making it into the first team. I asked Mike how many of the players made it to which he replied “one of the lads — from a squad of 30 — went on to get a 1st team cap for England”. A fantastic achievement. The journey of his rugby career is analogous to the start-up world — while raising a successful Series A round is an achievement, there is a long way to go from here…

We are going deep into data infrastructure

At MMC, we’ve been focused on AI and ML for over six years. Off the back of our research, we have built a large portfolio of incredible entrepreneurs leveraging the technology to do amazing things, including the likes of Peak AI, Synthesia, Signal and Qumata. One clear challenge to the adoption of AI in the enterprise remains the importance of high-quality data. We are creating 2.5 quintillion bytes of data every day (and growing exponentially). Not only is it expensive to store data, but the shelf-life is also shrinking.

We have been backing European entrepreneurs building across the data infrastructure stack for a number of years and have built a portfolio of companies that we are proud of that includes Snowplow, Ably, MindsDB, Tyk and Cloudsmith — and I passionately believe that Europe has the potential to create companies in this space that reach the scale of the US success stories in the sector like Snowflake and Datadog.

Quix has the potential to become a core component of every modern real-time data infrastructure stack.

Quix is a fully managed platform for business-critical real-time data applications with events processed “in motion”. It is the technology solution of choice for building and running event-driven data applications, and is already used by an impressive list of large clients such as Deloitte and McLaren. The Quix platform ties together nine separate technologies including Kafka, Kubernetes, Spark, and Git, to create an end-to-end infrastructure platform.

Input comes from multiple business event streams containing data such as customer orders, trading data, tweets, or sensor data from physical assets such as vehicles. The Quix platform processes the input data as it arrives, before optionally storing it in a persistent store. The platform is modular and configurable but abstracts the underlying infrastructure complexity from data teams looking to solve real-time analytics business use cases. It is built with Python, making it the obvious choice for data teams working with ML.

The platform empowers data teams, removing the need to involve engineering to enhance and maintain the real-time data infrastructure, and provides an ML ops solution for the data streams. Having spoken to numerous clients and potential clients the excitement about the platform is impressive!

We live in a real-time world.

We want things immediately: which stop the delivery driver is on and what time our package will arrive at home, the latest price of Ether, or the mileage remaining on our electric car. Enterprises are struggling with these challenges — consumers and businesses demand instant experiences across sectors such as media (news, interactive digital streaming), mobility (EV), e-commerce availability and corporate comms.

In addition, there are macro trends with broader cloud adoption, the proliferation of IoT devices, and the widespread implementation of 5G that are all driving an increase in the amount of streaming data available. More recently, the adoption of hybrid and multi-cloud environments is also creating further data connectivity challenges which in turn drives demand for real-time enabling technology.

“The world is on the brink of a real-time revolution in economics, as the quality and timeliness of information are transformed. “

Economist Cover story* Real-Time Revolution (October 23rd, 2021 edition)

We believe several market drivers position Quix well

— Real-time data consumption is growing and batch processing (the standard in many companies today) simply doesn’t cut it, especially with machine learning.

• Real-time is required to improve user experiences (ad ranking, Twitter’s trending hashtag ranking, Waze estimating your time of arrival) and to enable high-value event-driven applications (high-frequency trading, autonomous vehicles, voice assistants, fall detection for elderly care, fraud detection).
• Many large corporates and enterprises we speak to have started to experiment with an increasing amount of streaming infrastructure to enhance core use cases and unlock further value. When applying ML to any event-driven data stream having it processed in real-time delivers huge benefits which is a big part of the Quix value proposition.

— It remains complex to handle and process real-time event-driven data at scale.

• It is a time-intensive, complex task to build bespoke streaming solution and requires significant engineering commitment to monitor and maintain.
• In addition to streaming protocols, there are further complexities in managing software infrastructure — Kubernetes clusters, Devops, cost and performance monitoring all require effort. And while infrastructure management is a largely solved problem, infrastructure provisioning is still not efficient. This is a critical component for smaller companies looking for a broader data platform (v/s working on their own solutions at scale).

— Python is the natural coding language of ML

• Other data streaming tech such as Kafka and Flink run on Java and Scala. The Quix team chose to work in Python (v/s SQL for the broader market) — a language that supports deeper analytics, ML and application integration use cases, which differentiates it in the short-term and addresses the needs of data science teams that are used to working with Python.

We are excited to be backing the Quix team and our thesis can be simplified into three core elements:

1. The Quix proposition is aligned to the current direction of travel in data infrastructure

• Data scientists are looking for tools that empower them to become more hands-on and remove the complexity of handling stream data, the underlying infrastructure and the dependency on engineering for required changes. A growing number of streaming use cases require real-time insights. This has traditionally been achieved by feeding streaming data in low latency databases — still treating it as batch data. This is dealt with today by achieving transformations and analytics on the stream broker (like Kafka, or AWS Kinesis) itself which is complex and takes time. Quix solves this challenge.

2. Quix’s high-performance, scalable product solves real problems

• The product is technically impressive and customers validate the positive impact it can have. The roadmap and rate of product development positions it well vs alternatives and potential competitors. Some of the core Quix innovation sits in the libraries and packages for in-memory processing, protocols and helpers for non-expert users along with a choice of modular architecture helping teams avoid locking into a fixed protocol.

3. A mission-driven team

• The business was co-founded by a team of four former F1 engineers Mike, Tomáš, Peter and Patrick. Mike, CEO, was head of innovation at McLaren Applied and Tomáš , CTO, was responsible for developing the Telemetry Analytics Platform and ran a team that included Peter (Head of Platform) and Patrick (Head of Software).
• The team encountered the challenges Quix solves first hand at McLaren and we have been impressed with what they have achieved since founding the company.

The Quix team has a fantastic culture

We’re focused on a few core areas to help Mike and the team achieve the long-term potential of Quix:

• Building a repeatable, scalable GTM motion. Quix is still early in its journey and the team requires more commercial rockstars to help develop the commercial engine (reach-out if this sounds like a challenge for you!). There is encouraging momentum in the sales pipeline and the level of developer and data scientist engagement at conferences and events is amazing. Getting GTM is common for a company at this stage and something we spend a lot of time on with our portfolio.
• The market could take time to grow: Real-time stream processing is nascent and the market could develop in numerous ways, with large players like Confluent well positioned to own a big part of the market. The Quix team will need to remain agile and quickly iterate the product roadmap as the market matures.
• Product roadmap stacked with exciting features with some required to be fully ‘Enterprise ready.’ More powerful analytics — off-the-shelf transformations, data science / statistics capabilities (have a toolkit, but need more); Enterprise packaging — security, integrations with existing workflows (notifications, collaboration etc.) all planned to be built over the next 3–6 months. Once complete, the platform will become the stand-out solution for event-driven data applications.

Looking forward we’re excited about working with Mike, Thomas and the team as they look to rapidly scale Quix.

We spent a lot of time evolving the budget, a sensible activity particularly in the current macro climate, to ensure the company has a cash runway that should enable the team to hit some big milestones! Some of the things we will focus on include:

1. Refining the positioning of Quix: Getting the positioning right can take time and becomes something that is always being tweaked. With a platform that can do so much, it is often hard to develop messaging that resonates with all customer profiles.
2. Build out the commercial, partnership and product teams: The team is looking to hire several important product, developer relations, partnership and sales roles. Getting hiring right at this stage is critical and we’ll be doing all we can to help.
3. Scale the Quix Community: The team has developed an amazing following already and the momentum behind a bottom-up, developer-first community is an important element of longer-term success for Quix. The platform should become known across developer and data science communities as the best product in the market for event-driven data applications. We have seen first-hand the power of the Snowplow community and believe Quix has the potential to do the same here. If this sounds like a challenge you’d love to take on, the team is looking for developer advocates — if you’re interested, please reach out!
4. Board and advisors: We will continue to discuss the right structure for the Board as the company grows (this is an area where we have a lot of experience) but whatever the structure, getting some well-connected, experienced advisors around the business can make a big impact. We’re excited about opening up the MMC network.

Nitish Malhorta (my colleague and one of our resident data infrastructure experts) and I are delighted to be joining this adventure with our friends Project A and Passion Capital and look forward to supporting the team in delivering on this huge opportunity!

If you’re an entrepreneur building something in this area please reach out, we love learning about new ideas — ollie@mmc.vc

Follow Ollie on LinkedIn