Tag: Big Data

Our Investment in TwinStrand Biosciences, Leveraging Big Data And The Cloud To Improve Genome Sequencing Accuracy By 10,000x

Today, we’re excited to announce that Madrona has led the $16M Series A investment in TwinStrand Biosciences, a Seattle genomics company with the potential to profoundly impact all of us. TwinStrand’s technology will help detect cancer earlier when it can be most effectively treated, will help identify the most effective personalized therapies, and will help to recognize carcinogens quickly thereby lowering the development cost and time-to-market of powerful new drugs. We’ve previously discussed the incredible intersection of life sciences and computer science in our region – and TwinStrand is at the forefront of this amazing innovation opportunity.

When I first met Jesse, the founder and CEO of TwinStrand, he was discussing the technology in exclusively life science terms. However, as I listened, it was incredible how so many of the concepts had direct analogs to my experience with high scale software. TwinStrand’s “Duplex Sequencing” technology uniquely tags each strand of billions of individual DNA molecules with a chemical GUID. The DNA is then replicated to enable sequencing on a standard genome sequencer – resulting in up to 6 TB of data per run – then imported to the TwinStrand cloud where error correction algorithms are employed. The result is a high-resolution reading of the DNA sequence, 10,000x more accurate than standard sequencing. Duplex Sequencing reduces today’s DNA sequencing error rate of ~1% to below 0.0001%. This biochemical error correction approach reminded me of error correction techniques employed in high scale storage arrays in cloud datacenters.

Researchers are actively exploring how to use this level of precision to detect DNA mutations caused by chemicals (a market known as “genetic toxicology”). Today it can take more than 2 years to determine if a chemical is a carcinogen, as large tumors need to be given time to develop in lab animals. With Duplex Sequencing’s breakthrough accuracy, the resulting mutations can be detected as very small tumors within weeks – saving time, money, and the number of animals required. This testing is a critical step in the drug development process, but also is used to test the safety of agricultural chemicals, food contaminants, and even to examine the effect of space radiation.

When I talked with leaders in the clinical cancer community, a common response I heard was that this level of precision was amazing and insightful – but that today’s diagnostics don’t need that level of accuracy. This response reminded me of so many of the skeptics of 64-bit computing 15 years ago – who would ever need that much memory on any computer? With our investment, we are making the bet that new diagnostics, therapies and even information storage technologies will be developed to leverage this new precision, just like software has always found great new ways to leverage new system performance. It’s very exciting to see the future through the eyes of the TwinStrand team and invest in making it possible.

Jesse created Duplex Sequencing through his MD/PhD research with colleagues at the University of Washington. The TwinStrand team consists of half biochemists, and half software developers and bioinformaticians. Together, they have built an incredible foundation—contributing to more than 15 peer-reviewed scientific articles leveraging Duplex Sequencing and developing a portfolio of over 50 patents. To learn more, I’d suggest these three great recent articles:

TwinStrand’s product will be launching soon, and I look forward to seeing what scientists all over the world will create with it.

-Terry

P.S. Out of humility, Jesse doesn’t often share that he is the grandson of Jonas Salk, the scientist who discovered the vaccine for polio, definitively changing our world for the better. It’s pretty incredible to think that TwinStrand may have the same potential.

Innovate.AI – Announcing Our Investment in Envisagenics

Pictured S. Somasegar and Nagraj Kashyap

Today, in partnership with M12 (the newly renamed Microsoft Ventures), we are excited to announce the N. American winner of the Innovate.AI competition, a global startup competition to find the next generation of companies building intelligent applications.

The winning company, Envisagenics, has been awarded a $1M investment from Madrona Venture Group and M12, and we look forward to joining them on their journey to help solve some of the toughest problems in healthcare and biotechnology today.

Envisagenics Team.

We first explored the idea of an AI-focused startup competition with M12 in early 2017. Both Madrona and M12 have made large investments in AI, and we thought M12 would be the best partner to take this idea and help us reach a larger base of startups across North America. M12 was excited to partner with us on this and in addition took it global.

We received 250+ applications from companies in North America. Startups needed to have built products, achieved some level of customer progress and have raised limited capital to enter.

Working together with the team at M12, we pared this list of 250+ down to 9 finalists, who were working on intelligent applications in the fields of healthcare, financial services, cybersecurity, software development, enterprise sales, computer vision and manufacturing automation.

Each of the final teams were then invited to spend two days in Seattle at the Madrona and Microsoft offices to get to know one another and pitch to a team of VCs from both firms. The final presentations typically involved a presentation, Q&A, and live demo of the products, and we were very impressed with the quality of products that these teams have built.

Envisagenics, our winning company, is in the business of applying AI to RNA splicing. Founded by Maria Pineda and Martin Akerman, the company is using their proprietary data and machine learning models to build a drug discovery platform targeted at the large number of diseases caused by RNA splicing defects.

This is a big problem and an impactful solution. Approximately 15% of all diseases are caused by disrupted splicing, including 50% of rare genetic disorders. Healthcare companies can drastically accelerate their drug discovery processes by using Envisagenics’s data and machine learning models to prioritize drug targets and biomarkers.

We believe there is massive potential in the future of intelligent applications, and we are thrilled to be investing in a company that is applying the power of ML and AI to a key pain point in the world of healthcare.

Please join us in welcoming Envisagenics to the Madrona and M12 families!

Snowflake, a Cloud Native Data Warehouse and Our Newest Investment

Today we are announcing our investment in Snowflake. Snowflake is a cloud native data warehouse. Data warehouses have been used for years to store and analyze, not surprisingly, huge amounts of data. Over the past 5-10 years with the explosion of data and the rise of analytics & insights that this data provides, these stores have grown massively and are getting tougher and tougher to scale and manage in a cost-effective way. We are excited to back a company that embraced and leveraged the potential of cloud infrastructure from the start and is rapidly ramping their capabilities to meet the demands of enterprise cloud computing.

This investment is different from Madrona’s core strategy of investing at an early-stage in Pacific NW based companies. The company is later stage and is primarily based in Silicon Valley. But this company fits other Madrona criteria – the huge and growing secular shift to enterprise cloud computing, an A+ team with ties to Seattle and product and customer leadership in the emerging cloud data warehouse market. But even given this, why Snowflake?

Two of the massive computing trends we actively follow for investments are – the movement of enterprise computing and workloads to the cloud and the development of intelligent applications that make use of data through ML/AI and continuous learning. Both of these require and deal with massive amounts of data. For all the progress that we have made on these trends, we are still in the early phases of this tectonic computing shift – especially for enterprise customers. Many of the previous attempts to make enterprise applications available in the cloud have simply been a reworking of legacy applications, as opposed to cloud native design. We are seeing more technologies that are being designed and built ground-up to be cloud native. That’s exactly what Snowflake did for the world of data warehousing.

Benoit Dageville (co-Founder and CTO) and Thierry Cruanes (co-Founder and architect) came with a rich set of database experiences from Oracle, and they were joined by Bob Muglia as CEO in 2014. Bob is a very accomplished enterprise software and business leader, having spent more than 20 years at Microsoft including running Microsoft’s $16B Server and Tools business. Under Bob’s leadership, Microsoft grew several different multi-billion dollar businesses. Soma has had the opportunity to work for and with Bob over the years at Microsoft and everyone at Madrona sees Bob as a world-class leader. All this experience, expertise and background make Bob the ideal leader for Snowflake. We are really excited about this team and think they are the ones to create a meaningful new business in this industry.

Snowflake is a data warehouse designed and architected for the cloud. It is the first data warehouse built specifically to run in the cloud, and offers a range of performance, concurrency, scale and infrastructure management benefits which legacy, on-premises and cloud data platforms were not designed for. This allows Snowflake to achieve better database performance, respond to higher volumes of concurrent queries without performance degradation, and provide a simpler ongoing SaaS model without infrastructure maintenance – all with an outstanding price/performance characteristic.

Despite only being about 4 years into development, a recent GigaOm analyst report (http://info.snowflake.net/rs/252-RFO-227/images/GigaOm-sector-roadmap-cloud-analytic-databases-2017.pdf) ranked Snowflake as the top cloud analytics database ahead of Google BigQuery, Teradata, Azure Data Warehouse and AWS Redshift. While these other solutions can be a good fit in certain situations, we see Snowflake as a long-term leader in this massive market with its cloud-first technology and cross cloud platform potential.

Source: GigaOm

Snowflake is building a team in Bellevue given the cloud and big data talent that is available in this region. The combination of a world-class proven team, the focus on a cloud-native solution and the potential to be a leader in a massive cloud data warehousing and analytics market are the main reasons we decided to invest and participate in the Snowflake journey. Snowflake is built on Amazon Web Services (AWS) and there is a good partnership and collaboration between Snowflake and AWS. We look forward to being a valuable resource on that partnership given our long history working with AWS. In addition, we are excited to partner with Bob and team and help them build Snowflake’s presence here in the region and around the world.

Madrona Labs Adds CTO, long time entrepreneur and big data, machine learning, computer vision technologist, Jay Bartot

New Role Brings Deeper Level of Technology Expertise to Madrona Venture Labs Team
We are thrilled to announce that Jay Bartot has joined our Madrona Venture Labs team as Chief Technology Officer. Jay has been the cofounder of four successful startups, each of which leveraged big data and machine learning to provide targeted services to consumers and businesses. These startups include Farecast, where Jay and I partnered to change how consumers purchase airline tickets by using algorithms to predict airfare price fluctuations.

Joining Madrona Venture Labs is like a homecoming for Jay, as Farecast was incubated at Madrona Venture Group and he worked closely with Madrona’s Managing Director Matt McIlwain and Venture Partner Oren Etzioni in the earliest days of the company’s formation. At Farecast, Jay and I formed a highly productive engineering and product partnership and we aim to bring that same collaborative spirit to our Labs culture.

Jay’s other startups include AdRelevance, acquired by Media Metrix, Medify acquired by Alliance Health and most recently Vhoto, acquired by Hulu. Jay brings a wealth of deep technical and engineering leadership experience to our team and in support of our spinout founding teams. With Jay onboard, we will look to explore new, innovative technical startup ideas that leverage his experience in machine-learning and data-mining.

Jay is one of the most creative and inventive engineering leaders I know and we could not be more excited about our future with his influence and leadership.

Why We Invested in Integris.io

Today at Madrona we are excited to announce our seed investment in Integris, a data risk intelligence solution that enables companies to discover, classify and control how they’re using their customers’ data. This control enables companies to protect customer data and more easily comply with the complex set of privacy and consumer data regulations around the globe.

From our first conversations with Kristina, her vision resonated strongly and immediately with us. The ability to collect, store and analyze huge amounts of data from customers is immeasurably important to companies across every industry, but is being met with a wave of new (and frequently changing) regulations. The legal, customer-privacy, financial, and reputational risk of failing to meet this set of ever-evolving and complex regulations are dire, and companies experience significant challenges (and spend a lot of time, money, and resources) trying to keep up. It’s early days for Integris, but the team is building a powerful data risk intelligence solution that uniquely addresses this precise challenge for big companies in a proactive, automated, and continuous way.

We were so compelled by this team, the overall vision, and their progress, that we issued a term sheet before the company had even officially incorporated. When you encounter a team – and opportunity – like Integris’ and have the chance to back them even before “Day One”, you simply take it.

Following her stint as a highly respected Group PM at Microsoft, Integris’ CEO and co-founder Kristina Bergman became an all-star investor at Ignition, where she has been a principal for the past several years. On a personal note, there obviously aren’t too many women investors at institutional VC funds (much less in Seattle), so for years Kristina and I have always made a point of getting together at a regular cadence to exchange notes and ideas. I have profound respect for Kristina and for the way she both thinks and leads. Kristina spent late evenings and weekends over the past year and a half developing Integris, and it was energizing to spend time with her on the concept as it developed. It’s a delight today to officially partner with her as she and her talented co-founders – Uma Raghavan and Frank Martinez – bring Integris to life. We discussed with Kristina from the get-go the idea of creating a diverse syndicate for a seed round with complementary sets of experiences and networks; we’re very excited to partner up with Amplify, with Ignition, and such strong angel investors. It has been exciting to have our regular meetups evolve from two VCs chatting over interesting companies… to entrepreneur and board-member spending our time together on just one.

It’s just the beginning for Integris, but this is a world-class team taking a huge swing at a massive enterprise challenge. We’re all excited here at Madrona to get behind them.

 

How the Algorithm Economy and Containers are Changing the Way We Build and Deploy Apps Today

In the age of Big Data, algorithms give companies a competitive advantage. Today’s most important technology companies all have algorithmic intelligence built into the core of their product: Google Search, Facebook News Feed, Amazon’s and Netflix’s recommendation engines.

“Data is inherently dumb,” Peter Sondergaard, senior vice president at Gartner and global head of Research, said in The Internet of Things Will Give Rise To The Algorithm Economy. “It doesn’t actually do anything unless you know how to use it.”

Continue reading “How the Algorithm Economy and Containers are Changing the Way We Build and Deploy Apps Today”