Senior Data Scientist
Overview Charities accumulate large amounts of data about interactions with their supporters; but are generally ill equipped to use this information strategically. Accessible Intelligence has developed the most advanced application of machine learning to help charities use data to help them reach strategic marketing goals (e.g. increasing their monthly donation amount.) We are expanding our initial launch product to build more sophisticated models, and we are expanding our services to tackle a wider range of problems. The Role This data scientist role involves working alongside the existing team to further develop our machine-learning platform with the development of a new module for automatically generated text content for charity fundraisers. This position will largely involve the development of bespoke AI language models for our clients. These language models will be trained on our database of historical content combined with our unique content taxonomy. They will be used to generate example text in the voice of a given client based on prompts produced by our other systems. As this is the beginning of a new system for Accessible Intelligence the candidate will have some autonomy over development and architecture decisions such as the choice of frameworks, models, pipeline etc. As such a history in research level projects is highly desirable and a working knowledge of NLP and creation of bespoke language models or fine-tuning of large scale models (gpt-3 etc) is essential. Integration with and understanding of our wider product will also be important, the existing tech stack uses Python and SQL on top of AWS. These new models will be deployed and managed in AWS so familiarity with Linux, Docker, Git and cloud technologies would be good. In this project you will apply your machine learning knowledge to help improve the effectiveness of charity digital engagement. You will play a valuable role in the development of a cutting-edge software application. Responsibilities
- Developing bespoke language models for auto-generation of content for new and existing clients.
- Refactoring parts of our existing pipeline to ensure that it can integrate well with newly generated content.
- Apply the knowledge gained from these models to assist the content team in automatic classification of new content.
- Collaborating with the whole team on what is needed to help make the Accessible Intelligence vision into a reality through expanding the technology and products that we offer.
- Continuous development and evaluation of the platform to keep the platform up to date, efficient and secure.
- Specific knowledge of NLP and language model pipelines with experience working on at least one project in the area.
- A strong background in research and creating new software from the ground up.
- Experience with Python development and with the standard scientific libraries including, numpy, scipy, pandas.
- Knowledge of machine learning workflows in at least one common library/framework (e.g. scikit-learn, pytorch etc)
- Knowledge and experience using standard request APIs and SQL.
- Comfort with the basic Linux/Unix command line.
- Ideal: Experience in fine-tuning existing large language models (gtp-3 etc) to a specific purpose
- Ideal: Containerisation experience using Docker.
- Training and support provided
- Flexible working hours and locations
- Possibility to attend sector conferences to learn more and network
About Accessible Intelligence We are a UK start-up led by the CEO of another successful SaaS platform called Engaging Networks. The goal of Accessible Intelligence is to use artificial intelligence (AI) and machine learning to boost the effectiveness of charities’ supporter engagement. To Apply Email firstname.lastname@example.org with your CV and motivation (cover) letter. In addition, please state if you have the right to work in the UK. The company is looking for people with unrestricted and pre-existing rights to work in the UK. It is not registered to sponsor people for the job and is not looking for those on two-year graduate visas.