Scientist & Data / Machine Learning

PicturaBio; Packmoor; On-site; GBP 40,000 - 60,000; Full time

Job summary

A biotechnology company in Packmoor is seeking a Machine Learning Scientist to develop and maintain imaging-based classification models for rapid pathogen detection. The successful candidate will work closely with assay scientists and engineers, applying rigorous evaluation practices. A background in image-based data analysis and strong Python skills are essential. Candidates will thrive in a dynamic start-up culture focused on innovative solutions. This is an on-site role, ensuring close collaboration with laboratory teams.

Qualifications

• Strong experience with image-based datasets, especially fluorescence microscopy images.
• Proven experience in developing image segmentation and classification pipelines.
• Experience working with laboratory-generated or experimental data is an advantage.

Responsibilities

• Develop and maintain algorithms for pathogen classification in microscopy images.
• Train and evaluate machine learning models for biological particles.
• Work closely with lab scientists to interpret imaging data.

Skills

Strong Python skills

Scientific computing (NumPy, Pandas, SciPy)

Image processing and computer vision (e.g. OpenCV)

Deep learning for images (e.g. CNNs)

Version control and collaborative development (Git)

Clear communication skills

Education

Bachelor’s or Master’s degree in relevant field

Tools

PyTorch

TensorFlow

scikit-learn

Job description

Overview

Pictura Bio is a well-funded University of Oxford spin-out developing a novel diagnostic platform for rapid pathogen detection. Our technology combines fluorescence microscopy with automated image analysis and machine learning to identify pathogens in seconds. The platform is being translated into clinical diagnostic products, initially focused on respiratory infections, with broader applications across infectious disease.

Role Purpose

The Machine Learning Scientist will develop, evaluate, and maintain imaging‑based classification models that underpin Pictura Bio’s diagnostic platform. Working closely with assay scientists and engineers, you will analyse microscopy datasets, build robust and reproducible ML pipelines, and translate experimental data into validated diagnostic insights.

This role sits at the interface of data, biology, and regulated product development. We strongly prefer candidates who can work on site, collaborating closely with laboratory and engineering teams. You will be expected to work with real experimental data, apply rigorous evaluation practices, and clearly communicate results to both technical and non‑technical stakeholders.

Major Accountabilities

• Develop and maintain algorithms for segmentation, feature extraction, and classification of pathogens in fluorescence microscopy images
• Train and evaluate machine learning models for distinguishing viruses, bacteria, and other biological particles
• Perform data preprocessing, quality control, and exploratory analysis on microscopy datasets
• Work closely with lab scientists to interpret imaging data and feedback insights into assay and imaging design
• Build reproducible analysis and ML pipelines with appropriate documentation and version control
• Contribute to the integration of image analysis and ML models into production software
• Support performance evaluation using appropriate metrics, test datasets, and robustness checks
• Generate figures, reports, and summaries to support internal decision‑making and external communication
• Collaborate with software engineers to ensure models are maintainable, testable, and deployable
• Stay up to date with advances in image analysis, computer vision, and ML for microscopy and diagnostics
• Undertake other reasonable duties consistent with the role and level

Ideal Background

Education

• Bachelor’s or Master’s degree in Computer Science, Data Science, Machine Learning, Bioinformatics, Biomedical Engineering, Physics, or a related field
• PhD is desirable but not essential

Experience

• Experience working with image‑based datasets, ideally fluorescence microscopy images
• Experience developing image segmentation and/or classification pipelines
• Experience training ML models (e.g. PyTorch, TensorFlow, scikit‑learn)
• Experience contributing to software that is part of a product (e.g. deployed tools, internal platforms, or commercial software) is highly desirable
• Experience working with laboratory‑generated or experimental data is an advantage

Skills

• Strong Python skills and scientific computing (NumPy, Pandas, SciPy)
• Experience with image processing and computer vision (e.g. OpenCV, scikit‑image)
• Familiarity with deep learning for images (e.g. CNNs, U‑Net‑style segmentation)
• Ability to build reproducible, well‑documented data and ML pipelines
• Experience with version control and collaborative development (Git)
• Clear communication skills and ability to work effectively in interdisciplinary teams

Desirable Personal Attributes

• Comfortable working with messy, real‑world experimental data rather than curated benchmark datasets
• Pragmatic and outcome‑focused, with an interest in turning analysis into working product features
• Methodical and detail‑oriented, with a strong emphasis on reproducibility and robustness
• Able to balance research exploration with engineering discipline and deadlines
• Curious and willing to engage with wet‑lab scientists to understand data generation and experimental constraints
• Communicates clearly with both technical and non‑technical colleagues
• Enjoys working in a fast‑moving start‑up environment where priorities may evolve
• Proactive problem‑solver who is comfortable taking ownership of projects