Ethics & Standards
The standard we hold
ourselves to.
How we conduct engagements, handle client information, deploy AI systems, and manage conflicts — in plain language.
"We are not in the business of telling clients what they want to hear. We are in the business of telling them what they need to know — and building the architecture that acts on it. That commitment requires a standard we apply to ourselves before we apply it to anyone else."
Avilana Partners LLP
01
Professional Conduct
Honest.
Outcome-accountable.
Every Avilana engagement is delivered by the bench — not handed to a junior team after the pitch. The person who diagnoses the problem is the person who delivers the answer. This is not a positioning statement. It is how the firm is structurally built to operate.
We commit to honesty above comfort. If we believe a client's proposed direction is wrong, we say so — with evidence and a reasoned alternative. If we are not the right firm for a mandate, we say that too. We do not accept work we cannot execute well.
Engagements have defined deliverables and a defined exit. We do not extend scope to generate fees. We do not manufacture dependency. Our measure of success is whether the client no longer needs us for what we were engaged to do.
02
Confidentiality
What you tell us
stays with us.
All information shared in the course of an enquiry or engagement is treated as strictly confidential. We do not discuss client situations — named or anonymised — outside the engagement team. We do not use one client's experience to advise another without explicit permission.
We sign Non-Disclosure Agreements before substantive discussions in all engagements where sensitive business, financial, or strategic information will be shared. NDAs are mutual — we hold ourselves to the same obligation we ask of clients.
Confidentiality obligations do not expire at the end of an engagement. They extend indefinitely unless both parties agree otherwise in writing.
03
Conflict of Interest
Identified early.
Disclosed immediately.
Before accepting any engagement, we assess whether existing or prior client relationships create a conflict of interest. Where a conflict exists — or could reasonably appear to exist — we disclose it to the prospective client before proceeding.
We do not advise competing parties on the same or related matters simultaneously. Where a conflict arises mid-engagement, we notify the affected client immediately and determine the appropriate course of action together — including, where necessary, withdrawing from one or both engagements.
Members of the firm do not accept financial interests in client businesses as compensation for advisory work, unless this is explicitly agreed, documented, and disclosed as part of the engagement structure.
04
AI Ethics
AI that is accountable
to the humans it serves.
Avilana builds and operates AI systems — through Dragon Fire WorkStudio and the Einhar workforce division — that operate in real business environments with real consequences. We take that responsibility seriously and apply the following principles to every system we build or deploy.
Clients are told when they are interacting with an AI agent and what that agent is designed to do. We do not deploy AI systems that misrepresent themselves as human. The capabilities and limitations of every system we build are disclosed before deployment.
Every AI system we deploy operates under a defined governance structure with a human practitioner accountable for its outputs. Autonomous operation does not mean unaccountable operation. High-consequence decisions — financial commitments, client communications, regulatory filings — require human review before execution.
Client data processed through Avilana's AI infrastructure is kept in isolated, client-specific environments. It does not commingle with other clients' data. It does not train general models. It is not retained beyond the engagement unless explicitly agreed otherwise.
We review AI systems for outputs that could systematically disadvantage individuals or groups — particularly in hiring, credit, and operational decision-making contexts. Where bias is identified, we address it before deployment, not after.
Any party affected by an AI system Avilana has built or manages has the right to escalate a concern to a responsible member of the firm. Automated decisions that affect people's livelihoods, access, or rights are always subject to human review on request.
05
Client Data
Yours to access.
Ours to protect.
All data, documents, financials, and operational information shared by a client in the course of an engagement remains the client's property. Avilana uses it solely for the purposes of the engagement.
We retain client data only as long as the engagement or any post-engagement obligations require. Upon engagement closure, we provide clients with a copy of all work product and, on request, delete our copies of their raw data.
Client data is not monetised, packaged, or used to develop services for other parties. This is not a legal technicality — it is a condition of how we define trustworthy advisory.
"If you have a concern about how Avilana has conducted itself — with you or with anyone — write to us directly. We will respond, without intermediary, within 48 hours."