Services AI Products How We Work Case Studies About Trust Enquire

Trust & Transparency

AI assists the work. Alan signs off on it.

Kindred uses AI where it adds real value, and is open about when and how. Every deliverable is reviewed and approved by the director before it reaches a client.

AI in the work, on the record

Kindred is AI-agnostic and represents no vendor. The primary tool in active use is Claude, built by Anthropic, applied to four kinds of work.

01

Research and rapid analysis

Claude helps gather, summarise, and stress-test source material across large volumes of documents and data. It accelerates the reading, not the conclusions. The analysis and the judgment stay with Alan.

02

Drafting and document production

First drafts of reports, briefs, and proposals are produced faster with AI assistance. Every draft is then edited, fact-checked, and shaped by Alan before it is delivered.

03

Mapping and designing augmentation

Claude assists in mapping workflows and designing augmentation that removes manual, repetitive effort without removing the person doing the judgment work. The design choices are made against the client's context and constraints, not applied from a template.

04

Code and technical assistance

In technical engagements, AI assists with writing, reviewing, and explaining code. Alan tests and verifies anything that affects a client system before it is used.

Disclosure

Kindred has applied to Anthropic's Partner Program (applied June 2026, under review). This is the only vendor relationship of its kind. It does not change Kindred's AI-agnostic position or its commitment to recommend the right tool for each engagement.

Where the line sits

AI does not make professional judgments.

Recommendations, risk calls, and trade-offs are decided by Alan, not generated and passed on.

AI does not sign off the work.

Nothing reaches a client without being reviewed and approved by the director personally.

AI does not replace subject-matter expertise.

It supports work in areas Kindred already understands. It is not used to fake competence in areas it does not.

AI does not have the final word.

Where an AI output and Alan's professional view differ, the professional view governs the deliverable.

How your information is handled

Separated from AI-assisted work.

Confidential client information is kept in a separate, access-controlled environment from the AI-assisted drafting workspace. This separation is operational, not just a policy statement, and applies the same data classification discipline Kindred advises clients to adopt.

Not used to train AI models.

Where client information is processed with AI assistance, it is handled under Anthropic's commercial terms, under which content submitted through the API is not used to train Anthropic's models. Kindred does not place confidential client data into consumer AI products that may use it for training.

Used only for your engagement.

Client information is used solely for the engagement it relates to, under that engagement's confidentiality terms. It is not repurposed for other clients or other work.

Held only as long as needed.

Client material is retained for the engagement and a reasonable period after for professional record-keeping, then removed. It is not held indefinitely or built into a reusable data asset.

One name on every deliverable

Kindred Consulting is a one-director firm. Alan Taychouri is personally accountable for every deliverable that leaves it. There is no team to absorb a mistake and no firm's name to hide behind, which raises the standard rather than lowering it. Alan's background is in governance, risk, and compliance and in cyber risk, including a PCI DSS compliance uplift across all customer channels and a statewide, government-funded cybersecurity program. That experience shapes how AI is used here. The same data classification discipline he recommends to clients is applied to Kindred's own work, so confidential information stays protected at an operational level and accountability for AI-assisted output rests with one identifiable person.

01

Reviewed before delivery.

Every output is reviewed and signed off by Alan before it reaches a client. AI never delivers directly.

02

Disclosed when material.

Where AI has materially contributed to a deliverable, Kindred will tell you. You are never left guessing how the work was produced.

03

A commitment, not a checkbox.

This is a position Kindred chooses to hold, not a box ticked to satisfy a rule. It reflects how the work is actually done.

Read the full policy

The Kindred AI Use and Transparency Policy sets out in detail how AI is used, how client data is handled, and who is accountable. It is the document version of the commitments on this page, written for clients who need to see the governance in full.

Download the AI Policy (PDF)