SOFIE is hiring an Embedded Systems Software Engineer

SOFIE BIOSCIENCES, Inc., located in Culver City, California, is seeking an enthusiastic Embedded Systems Software Engineer to join a passionate team developing and supporting next generation macroscale and microscale radiochemistry systems.

SOFIE is a privately held, well-financed molecular imaging company. It is developing and offering a range of products from pre-clinical imaging systems and automated chemistry synthesis systems to new diagnostic imaging probes. More information on the company can be found at

The company’s environment will reward a self-starter — a creative problem solver who anticipates risks and opportunities and maintains a vision of continuous improvement. We work in a fun, collaborative environment that is as diverse, adventurous and open minded as the technology we develop. We encourage our employees to learn and grow personally and professionally. We have a culture that values great ideas by great people.

Seeking a strong, multi-talented candidate to join our dynamic, interdisciplinary team of scientists and engineers to develop disruptive technologies in Positron Emission Tomography (PET). In particular, this position will focus on the development of both embedded firmware and software for innovative, automated radiochemistry systems. You will be an integral part of a team making decisions and designing how users will interact with our automated chemistry platforms.

Bachelor’s degree in EE/CS or Computer Engineering + 5 years industry experience, alternatives accepted with relevant firmware, software, electrical, and UI development experience (E.g. – computer science, electrical engineering, etc.).

  • Experience with microcontrollers (ARM, Arduino or PIC), embedded computers and hardware integration
  • Knowledge and experience with electrical engineering fundamentals and circuits
  • Experience with PCB design, testing, and troubleshooting
  • Strong knowledge of C/C++ and Python
  • Have knowledge of at least one of the major python web frameworks, i.e. flask, pylons, cherrypy, django (flask preferred)
  • Experience with Git / Github
  • Familiarity with HTML5/CSS, JavaScript and web protocols
  • Ubuntu / Debian / Linux / Mac OS X including experience with terminal / command line
  • SQL database experience expected
  • Familiarity with RS232, RS485, SPI, I2C and TCP/IP communication protocols
  • Familiarity with client-server architecture and programming
  • Experience developing software architecture for automated systems and electromechanical components
  • Excellent written and oral communication, record keeping and organization, problem solving, and ability to work in a collaborative environment
  • Knowledge and experience in product development processes a plus
  • Legally authorized to work in the United States without company sponsorship

Job Duties

  • Develop firmware for embedded systems using C/C++ to communicate with sensors and electromechanical systems
  • Work closely with electrical engineers on PCB layout/testing and software to hardware interfaces
  • Development of the automated system backend software
    • Design and develop software architecture based on system requirements
    • Develop the server side application in Python and SQL (Flask + SQLite)
    • Python API to communicate with / control embedded systems
    • Maintain code base using software versioning (github)
    • Implement new features
    • Investigate and fix bug reports
  • Work closely with UX designers developing in HTML5/CSS/JavaScript
  • Develop system unit tests
  • Collaborate closely with an interdisciplinary R&D team to meet project milestones
  • Communicate ideas and information to scientists, engineers and non-technical personnel in both verbal and written form
  • Maintain thorough and detailed written records

Salary is competitive. The starting date is immediate for the right candidate.

To apply, send email cover letter with attached resume to:

Many probes, one box.

To facilitate a paradigm shift in the enablement of novel PET probes for the research and clinical community, a fundamental change in their routine synthesis methods is required. Standardizing the synthesis and purification through automation serves to reduce barriers to reliable, robust, safe synthesis, as well enable the ability to rapidly disseminate protocols across multiple sites.

Professor Michael van Dam has devoted the last decade of his multi-disciplinary career inventing technologies that can advance and accelerate research in cancer and other diseases. By bringing together physics, engineering, chemistry, computer science, and biotechnology, Dr. van Dam enables clinical research. Specifically, he focuses on creating tools for in vivo molecular imaging, including platforms to increase the diversity and availability of new positron emission tomography (PET) imaging probes, and platforms for molecular imaging of cells. They are also interested in advancing the technologies (e.g. microfluidics) frequently used in the lab.

One example of Dr. van Dam’s vision for PET can be seen in the ELIXYS Flex/Chem, originally invented and developed in his lab. We wanted to know more about this experience, so we met up with Dr. van Dam to discuss ELIXYS, the SOFIE Probe Network, and what is on the horizon for his research.

Melissa Moore: Tell us about the van Dam lab at UCLA’s Crump Institute of Molecular Imaging, and how the automation of radiosynthesis plays a part in your day-to-day research.

Mike van Dam: Our day-to-day research is automation of radiochemistry! We recognize the enormous value of PET in fundamental research, drug development, and medicine, and we see this value increasing as the number of different PET probes grows and enables monitoring of greater diversity of important biological processes. We also recognize the critical role that PET will play in transforming healthcare via precision medicine. To accelerate these developments, we want to put PET into the hands of more investigators. An important focus of our lab to develop novel technologies that can drive down the costs of producing PET tracers.

We also collaborate with several groups that are developing new PET probes or new basic chemistry methodologies, and automated technologies, including ELIXYS, have helped these investigators reduce radiation exposure and increase productivity.

As Director of the Crump Institute Cyclotron and Radiochemistry Technology Center, you also act as a service provider for the production of PET probes to pharma, biotech, and academia. What are some key challenges that you face in this role and how does ELIXYS help in these efforts?

One challenge we have faced is the diversity of PET probes requested by our customers. A lot of these requests have been for “established” probes that we did not have previous experience making, and thus we needed rapid and low-cost ways to establish the capability for reliably making these probes at UCLA. The ELIXYS has helped in a number of ways to streamline this process. The high degree of flexibility (3 reactors, wide temperature and pressure range) means we can typically use the reaction conditions as reported in literature, without ever running into limitations of the synthesizer that would require changes to be made. The R&D features allow us to monitor initial runs in detail to ensure that each step performs as expected in the system, and then instantly transition to routine production once adequate performance is achieved. These features have also been extremely helpful in projects to develop entirely novel tracers.

Another challenge we face is limited resources in the center. The Crump Institute radiochemistry facility is currently shared among 4 faculty research labs in addition to the probe production service, and contains only 4 hot cells and several mini cells. The ability of ELIXYS to make different tracers without any hardware reconfiguration has allowed our radiochemist to meet the diverse demands for probe production with only a single dedicated ELIXYS system (and single hot cell).

MM: You’re one of the first labs in the world to develop and produce as many probes as you have on a single radiosynthesizer in a single hot cell. Tell us about that journey.

MVD: The journey has only been possible because of the efforts of many talented students, postdocs, our radiochemist, and the contributions of colleagues at UCLA and other institutions.

It started during the development of the ELIXYS system itself. From the ground up, the ELIXYS system was designed to be extremely flexible so that it could synthesize nearly any probe. To convince the research community that we had achieved this design goal, we started by demonstrating that it was capable of synthesizing not only [18F]FDG (a straightforward probe often used to benchmark new technologies), but also probes like [18F]FAC (which required 3 reaction vessels as well as unusually high temperatures and pressures). We then developed protocols for a handful of additional well-known tracers (e.g. [18F]FLT, [18F]SFB, [18F]Fallypride) and several additional nucleoside analogs to demonstrate that tracers of varying complexity could be made without any need for system reconfiguration. At that point we were up to about 8 tracers.

Since then our efforts have been driven by our probe production service to meet the needs of various investigators. In the last year and half, we have added another 12 tracers to our list of “routine” probes, and we expect this to continue to grow. We have learned a few tricks for adapting probes onto ELIXYS, and this has helped speed our efforts when faced with the challenging of brining a new probe online.

It has really been a lot of fun for us to see how flexible the system actually is, and I’m really proud of the team that has worked so hard to get to this point!

MM: How do you see the SOFIE Probe Network enhancing your novel PET probe development?

MVD: Even though the ELIXYS synthesizer makes it very easy to adapt new synthesis protocols from the literature, it still takes effort to do this. At the very least, the newly developed ELIXYS program needs to be run a few times to assess the repeatability; in other cases, more significant optimization or changes may be needed before it is ready for routine production. It would certainly be much easier to directly download an ELIXYS program that has already been tested and optimized.

With a significant number of ELIXYS systems in use, the chances are increasing for any given probe that someone has already performed all of this work. The Sofie Probe Network will provide a very rapid and efficient tool for communicating these types of activities. We hope this is something the community will really get behind. For example, we plan to make available programs for all of the probes we have made so far so that others instantly have the capability to make all of these probes. We hope that others will do the same. We also anticipate the network will be used to share improvements and optimizations so the quality/repeatability of the protocols will be improving over time.

MM: As an inventor of ELIXYS, you have a clear understanding of its features. As a user, do you or your team have a favorite?

MVD: It’s hard to pick just one! The straightforward, drag-and-drop programming is very popular. This feature reduces the learning curve, making it easy for new lab members or collaborators to get started using the ELIXYS system for their research projects. For new probe development, the programming interface is a major time-saver, allowing us to create a bug-free program in a very short time and to be confident that it will perform the intended steps.

We also really like the features that support probe development and optimization. The ability to easily add pauses (i.e. “Prompt” unit operations) and access the contents of the reaction vessel is very beneficial when first starting to work with a new probe, or for optimization. We can remove the vial for accurate quantitation of radioactivity and sample the contents for radio-TLC or radio-HPLC analysis, providing detailed insight into each step of the synthesis. We’ll often take intermediate measurements at all stages so we can rank which steps to optimize first in terms of greatest potential for improving overall performance.

MM: What’s next for the van Dam lab?

MVD: In terms of the probe production service, we plan to continue ramping up our efforts to increase the diversity of available probes, and to increase our production throughput to serve more investigators.

In terms of our own research, we are focused on miniaturization and other strategies to further reduce the cost of PET probe synthesis. Considerable progress has been made toward an ultra-compact microfluidic synthesizer, and this technology is currently being commercialized by Sofie Biosciences. We are also exploring novel technologies, such as microscale purification and microscale quality control (QC) testing, to address remaining bottlenecks in the PET probe production process. As was the case with ELIXYS, we anticipate integrating these technologies into our probe production service to lower the cost of PET probes.

In addition, we are working with several collaborators for whom the ability to perform radiochemistry in small volumes provides fundamental advantages for advancing their research.

The long road to the clinic.

Positron Emission Tomography (PET) is a non-invasive imaging technology that uses a trace amount of a PET imaging agent (“probe”) to quantitatively and safely measure biochemical processes, including the cycle of disease initiation, progression, therapeutic intervention, and regression. PET is often referred to as “molecular imaging”; unlike clinical MRI and CT, which focus on structure, PET provides a real-time functional readout of the body.

One of PET’s defining characteristics is its translational nature – what can be studied in cells and mice can also be studied in humans – providing a powerful tool to evaluate the biology of disease and guide the drug discovery process; a diagnostic cornerstone of personalized medicine.

Professor Julie Sutcliffe knows the road to the clinic well. Her formative academic years as a chemist provided a window into PET probe discovery and routine clinical production. Now, as a tenured professor of Internal Medicine and Biomedical Engineering at UC Davis, co-Director of The Center for Molecular and Genomic Imaging, and Director of the new Radiochemistry Research and Training Facility, Dr. Sutcliffe’s research efforts are focused on the design, synthesis, in vitro identification, and in vivo screening of targeted molecular imaging agents. Dr. Sutcliffe’s group embraces innovative technologies, having developed and automated rapid radiolabeling technologies for the incorporation of fluorine-18 into peptides, becoming the first to publish the application of “click “ chemistry in PET and develop a high-throughput approach to PET probe discovery.

As the principal investigator for grants funded by the National Institutes of Health and the Department of Energy, and a track record of successful and productive research projects and education, she recently spearheaded the partnership between UC Davis Health System, PETNET Solutions Inc., a wholly owned subsidiary of Siemens Medical Solutions USA, Inc., and the Northern California PET Imaging Center (NCPIC) to establish a facility on UC Davis’ Sacramento campus for research and training in radiochemistry and for the commercial production of radiopharmaceutical products used in PET.

Few scientists are better suited to discuss PET’s translational power, so we sat down with Dr. Sutcliffe to find out about her latest Exploratory IND submission and plans for the future.

Melissa Moore: First, congratulations on your recent FDA submission for an exciting novel PET probe! Tell us a little about the process.

Julie Sutcliffe: Thank you, it was a marathon…painful but we survived! I would like to say it was fun but…(smiles).

MM: So, good, but good to be done?

JS: Yes, definitely.

MM: When did the idea for this probe first emerge? How did you know this would be a great candidate for the clinic?

JS: We became interested in the target αvβ6 approx 14 years ago. Many of you have heard about integrins and the RGD peptides to target integrin αvβ3. We had a lot of experience of making peptides – basically my life as a Ph.D student – but at that time there was nothing developed to target the integrin αvβ6. It was the new kid on the block, and was found to be overexpressed on head and neck cancers. I was working with a head and neck surgeon before I came to the USA and he described the challenges of achieving clean margins in these patients and the need for a way to better image the disease.

Actually, it is an epithelial-specific cell surface receptor that is undetectable in healthy adult epithelium but is significantly upregulated in a wide range of epithelial derived cancers. This receptor is often localized to the invasive front and infiltrating edges of tumors and plays a key role in invasion and metastasis and its expression is often associated with poor prognosis. With the unique expression of αvβ6 being a predictor of decreased progression free survival (PFS), response rate (RR), and overall survival (OS) we, and others, believe that the silver lining of this negative correlation is the much-needed opportunity to utilize αvβ6 for both diagnostic and therapeutic measures. The high contrast between malignant and healthy tissue and the functional relevance of αvβ6, especially in those diseases with a more aggressive phenotype, together place αvβ6 squarely on an elite list of targets for which development of diagnostic and therapeutic compounds will be vital to the future management of a very wide range of invasive diseases…so as it turns out it’s pretty important in many cancers, like pancreatic, lung, colon, breast, prostate and is an indicator for poor prognosis. In other words, you really don’t want your tumor to have αvβ6.

My first grant to the NIH (NCI) to develop peptide based molecular imaging agents to target αvβ6 was funded in 2003 by NCI, so off we set to develop the imaging agent. We used rational and random approaches as little was known about the target and its binding ligand (there was no crystal structure, for example) and believe it or not we found our answer in part of the foot and mouth disease virus! Lots of synthesis, lots of modifications, lots of long days and late nights in the lab, lots of fun and a few tears along the way and here we are αvβ6 -BP here we go! ☺

MM: That’s definitely how scientific discovery goes, right? A lot of hard work and the refusal to give up on good ideas. What was it that Edison said?

JS: Oh, yes! Something like, “I have not failed. I’ve just found ten thousand ways that won’t work”?

MM: That sounds right. So, you made it work! Can you share some key learning about preparing and submitting the IND package? What were the most challenging aspects of that process and how did you overcome them?

JS: Be prepared to get your hands dirty, especially if this is the first time your institution has done this. Be patient, it’s a lot of documentation. Don’t be afraid to ask for help. My friends at MSKCC, UCLA, and Wash U were lifesavers. Take off your research hat. Think about simplifying the chemistry early on so that you don’t have to make changes to the chemistry at the last minute to make it patient friendly. I’m hoping that the process is like riding a bike.

MM: Which means you never forget? Just get a little rusty in between?

JS: Yes! Hopefully the next one is even easier.

MM: I like to think of you as the Queen of Peptides. They’re definitely emerging as a powerful new class of PET probes and you were way ahead of the trend. What are some key features of peptides that make them great candidates for targeting disease?

JS: “Queen of peptides”, that’s a new one! I am usually referred to by many as “The princess”. But seriously, many peptide receptors are overexpressed in cancer, so we have a lot of targets and structures to choose from. Peptides are relatively easy to synthesize thanks to brilliant chemists such as Merrifield, the pioneer of solid-phase peptide synthesis. They are non-immunogenic and can be chemically modified to improve pharmacokinetics. They are also amenable to radiolabeling.

MM: Well, you wear many hats. Queen, princess – the entire royal court! Another amazing aspect of your lab is the way in which you embrace automation. You’re constantly forward thinking in your application of new tools to old problems. In what ways does automation affect your process?

JS: It’s definitely been helpful in removing operator error, improving reproducibility, and sparing the chemist an unnecessary radiation dose. Keeping my team safe and happy is critical.

MM: What role do you see the SOFIE Probe Network playing in the IND submission process?

JS: Currently there is no fully automated device to radiolabel peptides with fluorine-18. Method development and deployment to multiple sites is critical for multi-center trials. Sharing methods through the SPN would expedite development. The ability to cross reference tox and have rapid access to methods needed for IND applications would enable rapid deployment for multi-center trials. I’m excited to get going in the SPN!

MM: Me, too! What’s next for the Sutcliffe lab?

JS: We are excited to perform the first-in-human study. You’ve heard the phrase “bench to bedside and back again”, right? We expect to tweak a little more after we go into the clinic.

MM: Of course! The science is never quite done, right?

JS: Right. And if we are able to easily share methods and tox data, we will propose a multi-center trial. We have NIH funding to optimize the imaging agent further as well as develop the therapeutic partner for αvβ6. We also plan to use our combinatorial library screening strategies to identify new targets and develop new peptide probes. Not to mention a few graduations, a few weddings, a few bottles of champagne and having a LOT of FUN!