where we're headed

alice maz


systems biology is an exciting field quickly redefining the practice of biology. it provides tools and methods for researchers and bioengineers to model the behavior of biological systems. these models abstract over the details of molecular interactions. instead, they focus on accurately describing functional relationships between parts of the system, modeling the network as a whole. they are expressed as dynamical equations and can be simulated like any other physical model

however, running those simulations can be difficult. the dynamical properties of systems biology models aren't that complicated, but they aren't handled well by the standard differential equation solvers. they run much slower than they should have to, and a lot of error accumulates that could otherwise be avoided. accuracy is very important: simulation results may be used for anything from setting up lab experiments to justifying clinical trials

even worse, as models become large and complex, and especially when they need to be optimized, there's no suitable method to run them on a computing cluster. the immense labor required for scalable simulation causes many research teams and startups to avoid scaling their models at all. biologists themselves end up using generic tools like matlab and scipy for lack of more powerful and more specialized alternatives. and companies that offer automation and other services to biologists can't make use of these models at all

at sylph, we're attempting to solve these problems. we've developed a set of mathematical tools tailored to systems biology models. these tools provide direct access to scalable computing resources, significant reduction in error with formal proofs of error bounds, and an easy to use model optimization framework with drastically reduced simulation times. we're operating this as a cloud service, available to everyone at low cost. sylph provides a simple web api for model submission and evaluation along with clients that integrate easily with ipython and matlab. we believe this will significantly improve research practice in the field while lowering barriers to small teams and startups leveraging complex biological analysis

our long-term goal is to take this same strategy with hardware. we want to establish a highly automated process for design and execution of biological experiments. by standardizing a microfluidics experiment platform and generating chip layouts and control models from mathematical specifications, we can make developing a simulation and designing an experiment one in the same task. we want to produce a full toolchain for systems biology, bits to atoms, from idea to execution

client strategy

our initial customers come from the startup ecosystem. biology and biotech are rapidly growing markets, and startups drive the majority of the change. we provide broadly applicable tools for biology, so we're well-positioned to serve many of the big verticals. synthetic biology, drug discovery, genomics, lab automation, oncology, immunology, all these can benefit from what we're building

to demonstrate the value of sylph in practice, we'll consider its application in drug discovery, one of our most appealing target markets. it's already common to perform incremental simulation of the biological systems that a new drug targets. this is often performed with a collection of python tools that abstract away the underlying mathematics, tied together by a notebook interface. sylph can replace many of these tools and fit comfortably into existing interfaces while offering dramatically better results on several dimensions. we speed up simulation for a faster feedback-response loop. we increase the reliability of results. we slash the programming and mathematical skills required to test thousands of variations of a drug against its target. sylph fits right into a workflow that's already common, streamlining the process while increasing the confidence of the analysis

there's a lot of value to capture here. less than 10% of pharmaceutical r&d work makes it through clinical trials. the second phase alone, where drugs deemed safe for humans are trialed for efficacy, filters out more than half of substances that make it there. a candidate drug failing in clinical trials can cost a pharmaceutical company billions of dollars and could immediately kill a smaller company. companies are now adopting the systems biology approach to help manage these risks and improve their success rates in trials. we provide much stronger guarantees on the work these companies do, allowing them to shorten their development cycles and increase the percentage of drugs that make it to market

large pharmaceutical companies have notoriously difficult sales pipelines, but the landscape of pharmaceutical research has changed quite a lot in the last decade. more than half of pharmaceutical revenue is generated by "externally sourced products," drugs developed by small companies then licensed and pushed through trials by one of the large corporations. this tactic allows the pharmaceutical corporations to reduce r&d risk and allows research teams more flexibility than an internal team could have. given how resource limited these smaller companies tend to be, improvements in analysis and simulation tools can provide great benefit to them and dramatically decrease their risk exposure

sylph's applications are certainly not limited to pharmaceutical development. our methods have been developed for systems biology in its broadest scope. lab automation startups can run simulations to determine the optimal timing of operations and maximize useful throughput. synthetic biology companies can use our tools to rapidly test variations of gene sequences for optimal protein expression and gene circuit stability. oncology researchers can simulate the effect of a speculative treatment on a battery of cancer cell-line models, or conversely model a new cell-line and test its responsiveness to a wide variety of treatments. even a completely orthogonal research or engineering project can make use of these tools to support development of new methods in biology, as long as their mathematical approach is similar

finally, we want to establish deep roots in the systems biology research community. we have several research partners already excited about our work, and their feedback will be invaluable, especially as we refine our frontend experience for real workflows. but beyond that, we want to have a positive impact on the practice of systems biology. this is a very young field, especially in its mathematical foundations. we are making tools that serve its immediate needs, but we also want to help develop and standardize its methods and practice. ultimately we intend to serve as a bridge between academia and industry, setting up a virtuous cycle whereby our prospects rise along with the field as a whole

growth strategy

the cornerstone of growth is an unparalleled product, and we plan to win there on speed, reliability, scalability, and user experience. aside from that, our growth strategy is based on two principles: strong messaging, and ease of adoption

messaging for us means not only traditional marketing and branding, but also education and targeted outreach. there are tons of companies spread over a number of verticals that would benefit immensely from our platform, and most of them don't know it yet. for many of them, before we can lay out the specific advantages our software provides, we have to establish context for what kinds of things can be done with it. it's not enough for us to sell sylph on its technical merits. we have to sell systems biology

and... it's working sooner than we thought! since our last post here, we've started getting inbound customer emails. in all honesty we weren't really expecting this to happen. we'd started building sylph knowing the value of the methodology it supports, but we weren't sure how much convincing it would take to bring people on board with our vision. but based on the conversations we've had so far, founders understand the value our mathematical approaches can deliver to their companies. getting this level of market validation this early has of course been gratifying. going forward we'd like to keep a healthy mix of broad overview pieces, technical deep dives, and more directly educational material. our goal is to get on the radar of companies that have been waiting for something like sylph, while also bringing up to speed those who would gladly take advantage of us but aren't yet steeped in the methodology

we've been taking the opposite approach with the research community, reaching out directly to systems biologists to tell them about sylph. we're making a dramatically better replacement for the central piece of software in their workflow. explaining this is enough to garner interest. once we go live, I'd love to just give them accounts. research communities are extremely tight-knit, so if we win on product, the organic growth potential is huge. collaboration between institutions is everyday, top-down mentorship is basically mandatory, researchers transition between academia and industry and maintain ties to their colleagues on the other side. we help establish systems biology as a new wave in industry, then systems biologists bring us with them, wherever they go

hence the other guiding directive, ease of adoption. for organic growth to work, we need to tear down as many barriers as we can to adopting our software and our methodology. anyone can sign up to use our service, whether they're affiliated with an institution or not, and they can pay by credit card. of course we'll be doing a lot of heavy lifting for many of our early customers, considering how new our approach will be to most of them. but we want savvy startups, curious research groups, and lone biohackers to be able to get up and running on their own, like any good cloud service, from day one if they choose. then as we grow from a promising new approach to the standard platform for serious work, most of our onboarding gets picked up by the institutions that have reoriented around our existence

we also expect to make bottom-up sales in larger corporations once we've established our reputation among startups. a researcher or engineer hears about us through a friend, tries out our software on their own, then converts their team, then it spreads through the org. our tools don't need to interoperate with anything to generate meaningful and useful results, so the process can start with a single person. occasional small simulations can be run for minuscule cost, allowing us to get a toehold before our early adopters have to rope in procurement so to scale up to big models and fast workflows. ease of adoption and strong product win us users, coherent branding and marketing make us legible to purchasing managers. and if we have to negotiate contracts, we do it with the added leverage of already being in use inside the company. in this manner we can scale the business well into mid-stage before we need to consider the risks inherent in high-touch sales with large, unpredictable corporations

case study: twist bioscience

when looking for examples of similar strategies, it's impossible not to mention zoom. they're famously successful in driving bottom-up sales through the combination of strong product, widespread awareness, and few barriers to adoption. but personally, I like to compare us to twist

twist's core innovation is technical: a novel dna synthesis platform that outclasses all previous systems in accuracy, efficiency, and throughput. they compound on this strength by eliminating purchasing friction anywhere they can. they sell to anyone with a credit card directly on their site. they've cut minimum order sizes year after year, now as low as a few thousands bucks. their turnaround is a week or two depending on the product category. you can even order dna over their api

they tie this off with strong branding that makes them visible and legible even to those who don't necessarily understand their product. twist's sales funnel starts with someone wondering what a gene fragment library even is. their design and messaging communicate a feeling of centrality, like they're the focal point of their particular field. and, well, they are. they basically own the synthetic dna market, and their strategic plays now center on stoking new demand to meet their ability to supply it

overall they hold a commanding position, one that will reap huge rewards if they keep making progress on scale and manage to turn synthetic dna more into a routine consumable. we're leading with software in part to mitigate this kind of risk, of outgrowing the market while also shouldering a high burn rate. we build our customer base and establish market presence while operating comparatively lean, then we build out our microfluidics platform and the software feeds right into it. but in all respects, twist has put itself in an enviable position. they made a solid company serving a crucial need and they did it on the basis of product, messaging, and ease of adoption. they're the model I'd most like to follow


we're in the middle of raising our first round, and we're hoping to close soon. the intent of this post is to communicate our strategy to investors and customers who haven't reached out yet and to say we're coming to market as fast as we can. we've already started setting up deals to test our technology and mathematical approach in a lab setting, and we're going to expand this as far as we can go. we'll be launching our prototype early next year, which we expect will bring in even more interest. if you're interested in funding sylph or in getting early access to the platform for your lab or startup, reach out to me at alice@sylph.io. happy new year!