Source BioScience Limited is an international provider of integrated state of the art Laboratory Services and Products. Headquartered in the UK, with offices in UK, Europe and the USA, Source BioScience is a fully owned subsidiary of SourceBio International plc.

We’re looking for a Machine Learning (ML) engineer to assist with several on-going AI driven enhancements as well as to take ownership of the production of new models, frameworks and solutions within a rapid growth phase as the company integrates with hospitals across the country to deliver a next-gen digital pathology service. This remote role forms a core part of a development team within an exciting phase of rapid advancement and growth within the sector where Source LDPath is a leading driver of innovation and performance.

As part of the highly collaborative role, you will be working within a software development team and with our lead developer for machine learning/AI, where you will be collectively responsible for the design, implementation and deliverance of ML solutions within the digital pathology space of the company. There are currently a number of areas that are actively being developed and utilised within a production environment – the distinctions are in the area of Whole Slide Image (WSI) analysis (computer vision, classification, deep-learning) and within NLP utilising transformer-based models to utilise for NER, as well as utilising knowledge graphs, to extract information from input forms, diagnostic (and in the near future, potentially synoptic) reporting as well as other free flowing clinical text.  

The role also covers day-to-day involvement in other software engineering tasks for other aspects of deployed software as the team work across areas – which would include work on the new iteration of the company’s digital pathology flagship solution that has many layers of interconnectivity and interoperability including providing connectivity to-and-from hospital systems. This work can cover the entire life cycle of software development so the ideal candidate will need to be experienced across disciplines. The role will involve working on software development for the main digital pathology solutions and integration with them, as well as the specific machine learning needs of the job so any potential candidate will need to be comfortable working across disciplines – however, as growth continues, it is expected that the role will pivot towards the designing and implementation of ML solutions primarily.

The ideal candidate will be experienced and have worked as a machine learning engineer, software engineer or in a related role. The main environment is within AWS so experience of working with cloud platforms, especially within a machine learning or data science context, is very desirable.

The ideal candidate for the role should have experience of machine learning, data science, as well as DevOps and/or software engineering or development and will need to demonstrate practical experience. The role will require working within a team with rapidly growing responsibility and reach, so any candidate will need to be highly collaborative, an influential communicator and the capability to work on tight deadlines. Experience of working on digital health and/or pathology solutions is highly advantageous.

The role will require working or interacting on a range of non-ML solutions including modular software designed for the flagship system and related components, including but not limited to, lab management, reporting, integration, financial software connectivity, mobile applications, web applications, APIs both internal and external such as interfacing with NHS Trusts or nationwide interfaces or for specific clients/integrations.

The integration models work on a range of interoperability but primarily using HL7/FHIR, so any knowledge or experience in healthcare interoperability is advantageous, as it drives connectivity to send and receive clinical, diagnostic and digital information including digital slides.

This role involves pushing the boundaries of medical imaging and text AI within a high-volume context, real world settings, working with a large number of clients and internal tools to deliver impact across the business and to drive innovation in digital pathology.

About The Role:

The following qualities are what we are seeking for the role and how they would relate to their daily activities.

• Designing and implementation of ML based frameworks and solutions. The ML engineer would work with the lead developer for machine learning on designing and implementing required solutions. This process involves working with stakeholders to identify key requirements, deciding on key architectural and design decisions, development and testing, and then finally deployment within a production cloud environment. This would involve querying both large volume image data as well as other sources of data such as reports and clinical/diagnostic data extracted from messaging and databases (Postgres/SQL).
• Working on flagship software and modular components. The role will also involve work in more traditional software release settings initially to work on the company’s flagship software in digital pathology which involve work in a cloud-based environment potentially on both back-end and front-end components, as well as creation of, and interfacing with existing, APIs and other integrations.
• Data science and analytics. The role has capability for expansion into data science in terms of helping to create robust reporting and understanding across the area of the business such as helping to create analytical dashboards and query tools for front end users to drive context-driven understanding of the data.
• Working within an evolving CI/CD and responsive environment. The role will involve work in cloud-based environments, and by the nature of the software, will work on iterative releases.

The following are notes relating to required elements for potential candidates and description of ideal qualities.

• Effective, analytical and methodical designer. The role will require someone with a strong background in analytical thinking, techniques and the ability to break down complex problems. The ideal candidate will have a proven record of working on complex technical projects, utilising a range of techniques to potentially drive their thinking and approach to design, being able to integrate unique elements of healthcare and in understanding the needs of solutions within the space.
• Experienced data science, mathematical or big data skills. The role will require working with large, and complex data sets, including large volumes of free text and images, including a variety of formats and different levels of classification, metadata and related information. Candidates will need to be comfortable with dealing with a range of data sources, pre-and-post processing of data as well as strong understanding of analytical skills.
• Problem-solver. The development of any complex project can throw up a range of problems, potentially in a fast-paced environment with demanding deadlines. These problems can occur in the successful delivery of a project at any stage and with any stakeholders, so the engineer should be comfortable in helping to work through problems and to engage in removal of blockers and the implementations.
• Thorough and meticulous worker. The engineer will need to work across a range of software solutions and environments and will therefore need to be detailed and thorough in their working to ensure the highest levels of efficiency and safety while maintaining accurate records and collaboration within the development team – the team encompasses not just other engineers but will also interface with a business analyst, support staff, junior developers and with the leads for digital and cellular pathology within the business daily as well as a range of internal and external stakeholders.

As part of the development team, the main areas that the successful candidate will be involved in will be the developmental role where they will initially be working on:

• The flagship software (LIMS2) and related products
• On-going production machine learning models (NER from text data and image analysis of scanned slides relating to QC of slide quality)
• Working on designing new ML models, frameworks and solutions as required by the needs of the business and end-users

Ideal candidates should be comfortable working on both text and image based ML models, but experience is highly advantageous in either. The role will require working with, improving and implementing models across both domains.

Role Requirements:

Essential Criteria:

Candidates will be required to have the following:
 

• At least 3 years of experience working in a professional software development environment with experience in Python, C/C++ and C# and/or a Masters/PhD in a relevant field with good development skills
• Database/data management skills including SQL
• Experience of algorithm development, implementation and optimisation (while understanding/working to platform constraints)
• ML data acquisition and data management experience
• Experience in working with complex software and tech stacks
• Excellent problem-solving skills.
• Ability to work to deadlines.
• Self-disciplined and efficient, with a flexible and proactive nature.
• Excellent communication and collaboration skills.
• Strong workload and prioritisation planning skills.
• Ability to own and lead development from inception to deployment.
• Capable of translating high-level direction into design and implementation.
• Being able to work efficiently in a team environment as well as individually.
• Strong analytical skills
• Highly motivated and driven individual.

 Desirable Criteria:

These are criteria that will give candidates an edge, but they are not strictly required:

• Previous experience in healthcare software development/projects – especially with AWS
• Previous experience in deep and or reinforcement learning, including transformers
• Knowledge of healthcare interoperability, HL7/FHIR or any other protocols
• Working with large scale real-time systems, compression and optimisation
• Data science skills and analytics including backtesting, working with stochastic models in deep learning,
• Experience with classification modelling in digital pathology image analysis, CNNs, classifiers of bio image analysis and other related areas
• Clinical, diagnostic or healthcare related NLP experience

Benefits:

• Pension
• Income Protection
• Death in Service
• Employee Assistance Programme
• Medicash health cash plan
• Shopping & leisure discount portal
• Virtual GP Service
• Prove Yourself health and wellbeing resources
• SkinVision App
• Able Futures Mental Health support
• 33 days holiday (inclusive of bank holidays)
• An extra day off on your birthday
• Cycle to work scheme
• Discounted private healthcare

Candidate must have proof of their right to work in the UK.