Computer Vision Engineer
Posted on Friday, July 28, 2023
About Pear Bio
At Pear Bio, we are personalising cancer treatment selection because every cancer is unique. To achieve this, we’ve developed a test that cultures patient tumour samples and matched immune cells, quantifies dynamic cell behaviours during therapy exposure, and predicts treatment efficacy for that patient. The clinical, functional and multi-omics data that results from this process is used to identify patients with high unmet for target and drug discovery. We are a VC-backed start-up based in London. To grow our company, we’re looking for a Computer Vision Engineer to join our team. Will you be the one?
The computer vision engineer will work with Pear Bio’s R&D team to process high volumes of image data and create analysis pipelines to quantify biological behaviours that have predictive potential for helping cancer patients find effective treatments.
You will primarily work with time-lapse 3D microscope images taken of 3D cell cultures exposed to various therapeutics. Large volumes of data must be processed to create algorithms for object detection, classification and co-localisation. These image quantification methods will then be used to compare patients that are known to respond to treatment against those that resist treatment. You will create predictive models that will help guide future patients to effective therapies.
- Maintain a well-organised database of biological image data captured across numerous cell culture and treatment conditions
- Develop algorithms for object detection/segmentation and classification, especially based on the identity of different cell types and proteins
- Develop algorithms for measuring the colocalisation of cells/proteins
- Create object tracking algorithms for handling time-lapsed image data
- Create robust methods to handle varying magnifications and noise
- Create statistical and machine learning models to correlate image data to patient outcome data
- Find ways to accelerate image processing workflows through the use of cloud GPUs
- Maintain rigorous documentation of codebase
- Work with the Pear Bio lab team to determine the biological needs and relevance of each model
- Work with the software team to exchange methodologies/approaches applicable to different types of biological image data
- Bachelor’s degree in computer science, software engineering or biomedical engineering (with a focus on software)
- Highly skilled with Python, numpy, scikit-learn, pandas, matplotlib, OpenCV
- Experience with processing image data, especially analysing and manipulating pixels/voxels or pixel intensity equalisation
- Experience implementing the latest machine learning and/or deep learning techniques for high volumes of image data in classification and prediction models
- Master’s or 2+ years of work experience in computer vision and/or machine learning
- Experience using cloud GPUs for accelerated image processing and machine learning applications
- Understanding of traditional Machine Learning tools such as KNNs, K-means, SVM, decision trees, etc
What’s in It for You
- London office/lab space
- Competitive compensation in line with industry standardsStock options in a growing start-up
- 28 days of annual leave excluding bank holidays and Christmas closure
- Yearly personal development budget, plus the chance to represent the company at international conferences
- Open work environment where your opinions are valued
- High career growth & personal development in a fast-paced, dynamic environment
- Company perks/discounts via Perks at Work
- The chance to have an impact in shaping the future of an early-stage start-up
- We are unable to sponsor work visas at this time. Please confirm your ability to work in the UK without visa sponsorship before applying.
- The position is not eligible for remote work, so you’ll be expected to be on-site
We are unable to sponsor work visas at this time. Please confirm your ability to work in the UK without visa sponsorship before applying.
The position is not eligible for remote work, so you’ll be expected to be on-site.