Deepcell

11 - 25 Employees
< 10 Engineers
$10M - $25M Funding
Series A

Today, most blood-based genetic tests are exploratory and typically followed by an intrusive real diagnostic test. Our blood, however, is a treasure trove of information if handled properly. Truly diagnostic blood tests could dramatically change the way we manage our health personally and our healthcare as a society.

Recent advances in various areas of science and technology have created a possibility for truly diagnostic blood tests. Deepcell is developing a disruptive technology based on AI/ML, microfluidics and genetics that offers a non-invasive blood test with diagnostic-level accuracy.

Deepcell photo 1 Deepcell photo 2
Active Roles
Why join us?
  • Solve challenging problems that could revolutionize healthcare.

  • Funded and backed by top tier VCs in the valley.

  • Won multiple prestigious awards.

  • Founded by ex-Googlers, Stanford Alums. Team background from Stanford, Caltech, UCLA, USC, UCSF, Google, Microsoft Research, Roche, Invitae, JPL, Boeing, Fluidigm, Pacific Bio, etc.

  • Work closely with experts in biology, hardware design, and bioinformatics.


Engineering at Deepcell
Engineering team and processes

We have a rapidly growing and driven engineering team which is split between software, hardware, and mechanical engineering.

Software engineering follows 2 tracks:

— Instrument software is written in C# and C++.

— Data engineering uses Python, Tensorflow, and Apache Beam.

We operate on a weekly cycle. Every Monday, we review the tasks and discuss the action items for the week. The bigger goals (aka OKRs) are discussed and decided with the team on a quarterly basis. We do code reviews seriously. We use Jira to track bugs and tasks.

Technical Challenges

We deal with a huge amount of data and it is growing at an incredible pace. We are already in the range of hundreds of millions of data points! This data contains an incredible amount of information, which needs systems and processes to manage and extract.

We are working with “bits”, “atoms”, and humans. Our system parameters often require handling and processing events in realtime (with latencies in millisecond). We are constantly looking to push the system to increase its throughput. And we do all of that while striving to design the best human experience, be it the physician, the patient, or the machine operator.

Projects you might work on
  • Parsing analog input sampled at tens of thousands of samples per second in an efficient manner.

  • Controlling hardware with millisecond accuracy and improving its throughput.

  • Image processing to auto-adjust parameters in the instrument.

  • Building a beautiful, intuitive UI to run the instrument.

  • Connecting the dots between samples, instrument runs, and downstream biological results.

  • Extracting information from hundreds of millions of data points.

  • Train models that need to achieve F1 scores of 99.99%.

  • Asynchronously control multiple hardware components to orchestrate instrument runs.

  • Refactor code to simultaneously improve readability, testability, flexibility, and design.

Tech stack
python
Tensorflow
C#
Google Cloud
C++

Working at Deepcell

In one word, we treasure cautious optimist attitude. Optimism alone may result in too much disappointment and too much caution does not get anything done.

Free Food

We use EatClub!

Beautiful Office

We just renovated the office. We are near a Google office building in Mountain View.

Gym/Fitness

We cover gym membership.

Health Insurance

We provide medical, dental and vision.

Workshops/Conferences

We encourage participations in relevant workshops and conferences.


External Links

Interested in this company?
Skip straight to final-round interviews by applying through Triplebyte.

Apply