1. What is ReaDI-Watch?

ReaDI-Watch is a structured RD&I platform backed by an expert team. We help organisations define, capture, and manage their R&D and Innovation (RD&I) activity in a consistent way. So it is organised, measurable and ready for funding, reporting and decision-making.

2. What problem does ReaDI-Watch solve?

Most companies innovate continuously, but the development journey is scattered across project management tools, meetings, emails, spreadsheets, and memory. When it comes time to submit an R&D Tax Credit claim, this issue becomes a huge burden. ReaDI-Watch brings that activity into one structured system. Projects are clearly defined. Progress is captured properly. Time and cost are aligned. Leadership has visibility. Nothing needs to be rebuilt at year-end.

3. Is ReaDI-Watch software or a service?

It is both. Our platform provides the structured environment. Our team runs the process, defining projects, maintaining cadence, structuring updates, and supporting reporting. You don’t just get access to a tool. You get a managed system.

4. What changes after working with ReaDI-Watch?

RD&I stops being informal and reactive. Projects are framed clearly from the outset. Documentation is captured as work happens. Financial alignment is cleaner. Leadership has clarity. Instead of scrambling to explain RD&I later, you are building a structured innovation record in real time.

Ai Policy | Readi Watch

Data Management

Ai

Ai Policy

Our Approach to AI

AI is central to what we build and how we work. This page explains what we use it for, what principles govern its use, and what that means for your data.

This statement is reviewed annually and updated whenever we make material changes to the AI tools we use or the principles that govern them. Last updated June 2026.

Our Position

We have spent years working with companies and applied research centres on R&D across multiple industries and jurisdictions.

Along this journey, we see the same two problems appear everywhere, regardless of sector or scale:

Organisations struggle to define, measure and manage their R&D consistently. There are few common standards, shared frameworks, and common language for what R&D actually is in an industrial and applied context.
The difficulty in defining, measuring and managing R&D creates a deep struggle to capture new R&D knowledge generated within the organisation. The experimental journey, the decisions taken, the knowledge generated, tends to remain tacitly in people’s memory. It tends to not be recorded in a form that can be reused, built upon, or defended.

ReaDI-Watch addresses both of these problems. It provides the framework that has been missing, a structured way to define, govern, and capture R&D as it happens. AI is transformative in making that practical. It makes rigorous, real-time knowledge capture fast enough and natural enough that it actually gets done. But it does not replace human judgement. The decisions, the validation, the authority over what gets recorded, these remain with the people doing the work. That is how we have designed it, and that is what makes the output trustworthy.

Five Principles for AI

Every AI design and development decision for the ReaDI platform is guided by five principles. They apply equally to how we deliver services to clients today and to how the platform itself is built.

1. Human governance — AI assists, humans decide.

AI sequences tasks, structures outputs, and surfaces knowledge. It does not make decisions on your behalf, resolve uncertainty independently, or overwrite records without instruction. Authority remains with the user at all times. Nothing the AI generates becomes part of your knowledge base until a named person has reviewed and approved it.

2. Temporal integrity — the record cannot be rewritten.

Approved records are immutable. Once content is sealed, its substance, sources, contributor attribution, and approval timestamps are locked and cannot be modified — not by users, not by administrators, not by the AI. Later understanding is captured through new records linked to their predecessors, not by revising earlier ones. This creates a development narrative that is authentic and defensible over time.

3. Progressive validation — only approved knowledge becomes the foundation for future outputs.

The platform builds knowledge progressively. Each approved record draws on validated records before it and becomes the basis for what comes next. Unapproved or weakly supported content is flagged and cannot propagate through the system. Accuracy compounds over time rather than errors compounding.

4. Constrained autonomy — agents support progression but do not independently determine outcomes.

When the AI encounters uncertainty or incomplete information it has three permitted behaviours: proceed where sufficient information exists, flag what is uncertain or weakly supported, and ask targeted questions to resolve material gaps. It does not fill gaps with assumptions or present inferred content as factual. Users can always see what the AI is confident about and what it is not.

5. Transparency — users always understand what the AI is drawing from and how confident it is.

Before any AI-assisted session begins, users see exactly which source material has been selected and can adjust it. During generation, each output carries a confidence indicator showing how closely it is grounded in that source material. Where confidence is low, the flag identifies what is weakly supported and why. Every override decision is recorded permanently.

AI in Our Work

AI is already part of how we deliver services to clients today — in research, analysis, knowledge structuring, and documentation. The same five principles apply to our service delivery as to the platform itself.

When a ReaDI-Watch team member uses AI to support a client engagement, the output is reviewed, validated, and approved by a qualified person before the client sees it. The AI accelerates the work. The human is responsible for it.

We also use AI internally for operational tasks including drafting communications, summarising research, and preparing meeting notes. The same rules apply: review before use, no client data without a documented basis, no unapproved tools.

When we use AI transcription on calls, we tell you at the start of the session. You can decline, and we will accommodate that.

The AI Tools We Use

Provider

What we use it for

Location

Tactiq

AI language model powering the ReaDI platform; internal drafting and analysis

Supplementary AI language model capability

Meeting transcription and summarisation

United States

Anthropic (Claude)

OpenAI

All three providers are listed as Subprocessors in our Privacy Policy. Each operates under enterprise agreements that include explicit commitments against using your data to train their models.

Your Data

When you use the ReaDI platform, your data is processed by AI to help generate reports and surface knowledge. That processing happens within the architecture described above: permission-bounded, human-reviewed, and never used to train models.

We apply the following rules without exception:

Your data is processed only to deliver the service you have asked for.
It is not shared with other organisations or used to inform outputs for other clients.
It does not leave the ReaDI infrastructure except through scoped prompts sent to our AI providers, which are subject to their enterprise data protection commitments.
Sensitive financial data — including salary information, cost structures, and funding projections — is accessible only to users with explicit permission. That boundary applies to the AI as well.

Applicable Laws and Regulations

ReaDI-Watch operates across multiple jurisdictions and designs its AI use to comply with the data protection and AI governance frameworks applicable in those environments.

General Data Protection Regulation (GDPR) — EU/EEA

Where ReaDI-Watch processes personal data on behalf of EU or EEA-based clients or their employees, we act as a data processor under GDPR. Our clients act as data controllers. The lawful basis for processing is the performance of a contract (Article 6(1)(b)) and, where applicable, legitimate interests (Article 6(1)(f)). Data subject rights — including access, rectification, and erasure — are governed by the terms of our Data Processing Agreement and Privacy Policy.

PIPEDA and Law 25 — Canada

For Canadian clients and engagements, ReaDI-Watch operates in accordance with the Personal Information Protection and Electronic Documents Act (PIPEDA) at the federal level, and Québec’s Law 25 where applicable. Our data handling practices are designed to meet the requirements of both.

EU AI Act

The EU AI Act is now in force with obligations coming into effect on a phased timeline through 2027. ReaDI-Watch is monitoring its applicability to our platform and service delivery. The constrained autonomy model at the core of ReaDI — which requires human review and approval before any AI-generated content enters a client’s knowledge base — is designed to align with the transparency and human oversight principles the Act requires. We will update this section as regulatory guidance matures.

For full details of how personal and company data is collected, stored, protected, and subject to individual rights, see our Privacy Policy and Data Processing Agreement at readi-watch.com/privacy.

Enterprise and Sovereign Deployments

For larger organisations with specific data residency, sovereignty, or regulatory requirements, the ReaDI platform is designed to accommodate deployment models that go beyond our standard configuration.

Depending on your organisation’s requirements, we can explore:

Private cloud deployments where your data is processed on dedicated infrastructure rather than shared endpoints.
Self-hosted open source models deployed entirely within your own infrastructure, where no data crosses your organisational boundary at any point.
Hybrid arrangements where the knowledge base and retrieval layer sit inside your environment and AI processing is handled through a private, contracted endpoint.

If this applies to your organisation, get in touch to discuss what is possible.

Questions

If you have questions about how we use AI, how your data is processed, or what protections apply to your organisation, we are happy to discuss it directly.