AI Statement
Responsible AI on the SimplyDo Platform
Last updated: July 2026
Whilst recognising the huge potential and advantages of generative artificial intelligence via the implementation of
LLMs (large language models), and its application within SimplyDo’s platform technology, our priority remains the
safeguarding of our customer data and ensuring that we deliver technology responsibly and securely. We understand that
some customers are more resistant to adopting AI due to uncertainties in its safety, accuracy, security, or for any
other number of reasons.
This document describes the key areas of SimplyDo’s approach to developing and deploying artificial intelligence
technologies as part of its solution suite, and aims to help our customers understand the choices they have and how we
provide such capabilities safely.
Optionality
The SimplyDo platform is, and will remain, fully usable without enabling or using features based on AI. Nearly all AI
features are developed to enhance existing workflows through offering time-saving enhancements or scalability
opportunities rather than introducing entirely new feature sets.
As such, we currently consider our AI features to mostly be progressively-AI (rather than AI-first), with alternative
workflows maintained or developed in order to achieve the same or similar results using “traditional” methods. This
allows customers to elect to use such capabilities at an organisational level or to sporadically use them on a
needs-basis (e.g. for a specific challenge).
Customer account managers will assist in the understanding of these features, and will guide choices depending on
customer need.
Utility
SimplyDo’s current and future development and deployment of AI-dependent capabilities are based purely on usefulness and
where there is an opportunity to deliver real customer value. Whether this is in the development of internal tooling for
our staff to work more scalably with our customers or directly in user-facing features, we take care to ensure that new
technologies are not simply shoehorned in just for the sake of it or as a marketing tactic.
This means that customers can be sure that when SimplyDo offers AI-based capabilities, this is done with a real
intention to enable customers to save time, become more efficient, scale their workloads, experiment, or provide other
useful outcomes.
Data Sovereignty
We strongly believe that our customers should remain in control of their data at all times. This has been a pillar of
our approach to data protection for a number of years, with processes and controls in place at a technological and
organisational level to ensure that customer rights over their own data are fully protected. This includes a strong
understanding of foreground and background IP (data ownership boundaries), data segregation (enforcing separation of
concerns between customers and deployments), data access (through APIs, exports, and reporting), and data deletion
(through software controls and account management).
This stance extends to all features, including those based on AI, and we ensure that the same rights and controls remain
in-place for customers using these features. This allows customers to depend on SimplyDo to ensure security and control
over how (and where) the data is processed and stored.
SimplyDo platform customer data has been processed and stored fully within the UK since 2018, and this includes all
backups and redundant systems. We strongly restrict and carefully audit any new data processor suppliers (data
“subprocessors”) and ensure that data they process on our behalf is kept within the UK. We currently use the UK data
centres of the following suppliers for processing or storing customer data, including for AI workloads:
- Amazon Web Services;
- Microsoft Azure.
Model Training & Inference
Training is the process of creating a new model based on a large training dataset. In machine learning (and AI by
extension), a trained model can then make predictions (or inferences) against future known or unknown inputs. Some
companies use a mix of broadly or publicly available data and/or customer data in order to train their models, the
latter of which can risk exposure of the customer data if the model is then used by external users.
SimplyDo takes care to ensure that customer data is never used for training AI models. All SimplyDo AI features are
built on existing established models and we do not build or train our own. Furthermore, we only ever make use of AI
services where there is a guarantee that customer input data to trained models (i.e. “prompts”) are not logged or
permanently stored (beyond caching) or used to train subsequent models.
As with all SimplyDo technology, care is taken to apply industry standard data protections and encryption to protect
customer data in-transit, at-rest and, where possible, during processing.
SimplyDo AI features are currently built leveraging Amazon’s (AWS) Bedrock service. The following resources offer
helpful information to those wishing to find out more about how these services are delivered:
- AWS Bedrock Security and Privacy: https://aws.amazon.com/bedrock/security-compliance
- AWS Bedrock Data Protection: https://docs.aws.amazon.com/bedrock/latest/userguide/data-protection.html
Governance
SimplyDo maintains and regularly audits an Artificial Intelligence Register, which enumerates and details the AI
capabilities and underlying services in-use by our platform. This includes the features, how they are controlled or
restricted, models, services, and any related subprocessors.
The Register is subject to our established change management mechanism, meaning that any changes must be approved at
board level and by our Data Protection Officer before they are introduced to our platform.
For any questions about this statement or any other topic related to SimplyDo’s use of AI, please contact
support@simplydo.co.uk, and your query will be routed to our technology and data protection leads.
