Truveil is onboarding its first beta cohort. Request access

Make every AI agent decision defensible.

Truveil captures every action your AI agent takes, scores it against the regulations that apply in your jurisdiction, and produces an audit report regulators and procurement teams can act on. Plus free regulatory advisory in Claude, ChatGPT, and any MCP-compatible tool.

Six regulatory frameworks Proprietary scoring engine Patent pending Used in production

Your agent finished the task. What did it decide along the way?

Somewhere in that run it chose what to act on alone, what nobody can undo, what no one reviewed. Truveil surfaces every decision, names the gaps, and gives you the fixes. Not a smarter agent. A safer one.

What Conduct Assurance means →

Meet Truveil where you already work.

Truveil works inside the tools you already use. Ask in Claude. Ask in ChatGPT. Add to Claude, ChatGPT, Cursor, Make, Zapier, or any MCP-compatible client. Using n8n? Install the dedicated community node. No new dashboard. No behaviour change. Just regulatory expertise where you already think.

Five ways in.

From a one-line MCP connection to full production instrumentation. Every path feeds the same tamper-evident trail.

MCP Connector

Claude, ChatGPT, Cursor, Make, Zapier, or any MCP-compatible client. Connect in minutes, no code. Endpoint issued after signup.

See the Integrate page for setup.
Node SDK

One dependency-free file. The full capture contract: every accountability signal, one named line each.

See the Integrate page for setup.
Python SDK

The same contract, ported method for method and enforced by the same tests. Nothing lost in translation.

See the Integrate page for setup.
HTTP API

Any language, any stack. The SDKs are thin wrappers over this surface; build against it directly.

See the Integrate page for setup.
n8n Community Node

A dedicated node in n8n's native format. Install, add your API key as a credential, drop Truveil steps into any workflow.

See the Integrate page for setup.
Five integration paths. One tamper-evident trail. Zero behaviour change.

Who answers when your AI agent is wrong?

When an autonomous agent makes a consequential decision (hiring, credit, content moderation, medical triage), someone is legally responsible. Truveil makes sure it is not a mystery.

An agent decision nobody can explain is a liability.

Every action your agent takes is a decision someone may have to account for. Without a structured record mapped to the signals regulators actually check, you cannot explain it to a customer, an auditor, or a court. The record has to exist before you need it.

Enforcement is shifting, not disappearing.

The Digital Omnibus deferred the EU's high-risk obligations to December 2027 for Annex III systems and August 2028 for Annex I products. India's DPDP audit obligations bind from 13 May 2027. UAE's Federal PDPL executive regulations are pending. The grace period is real. The eventual obligation is not.

Your buyers are asking before your regulator does.

Enterprise procurement teams already ask for AI accountability evidence, and ISO 42001 is becoming RFP table stakes. An agent you cannot describe in their terms is a deal you do not close. Truveil produces what they ask for, from the run you already logged.

Truveil is built so the audit trail you produce today is the one regulators ask for tomorrow. Start before the deadline, not after.

Your AI agent makes hundreds of decisions a day. Most teams cannot prove which ones were reviewed, which were reversed, and which never should have happened. Truveil produces the evidence trail your regulators, auditors, and clients will eventually ask for.

From agent log to audit report in four steps.

Works with your existing agent stack. No new infrastructure required.

01
Integrate

Connect your agent in minutes: attach the MCP to Claude, ChatGPT, or any compatible client, or drop the SDK into your agent's codebase. Works with what you already have.

02
Capture

Every decision, every action, every approval flowing through your agent is captured as structured evidence in real time.

03
Score

Truveil's proprietary engine evaluates the evidence against the regulations that apply in your jurisdiction. Calibrated against six major frameworks. Category-aware so a research assistant is not graded like a hiring AI.

04
Report

Audit-grade reports in under 30 seconds. Plain language for legal and procurement. Regulatory citations for defence. Tracked over time so you can prove improvement.

Six frameworks. One audit engine.

Truveil's engine knows what each framework demands. Your audit report cites the right one for your jurisdiction. Cross-jurisdictional deployments scored against the strictest applicable rule.

EU

EU AI Act

Articles 5, 9-15, 22, 25-27, 49, 72-73. Annex III high-risk categories. Article 14 human oversight. FRIA scope.

US

NIST AI Risk Management Framework

Govern, Map, Measure, Manage functions. Generative AI profile. Sectoral overlays for credit, employment, clinical, housing.

IN

India DPDP Rules 2025

Rules 3, 6, 8, 9, 13, 14. Significant Data Fiduciary obligations. Rule 13 DPIA and audit cadence. Rule 6(1)(e) log retention.

DIFC

DIFC Regulation 10

Autonomous Systems Officer where high-risk processing applies. Evidence of compliance on request under Regulation 10.2.2. Significant-decision oversight. Certification pathway.

ISO

ISO/IEC 42001

AI management system standard. Impact assessment, operational control, continual improvement. Certification-ready evidence.

SG

Singapore Model AI Governance Framework for Generative AI (MGF-GenAI)

Nine dimensions of accountability. AI Verify alignment. MGF-GenAI profile. IMDA expectations.

Built for every team that ships AI.

Whether you are writing the code, signing the contract, reviewing the vendor, or reporting to the board, Truveil gives you what you need.

Developers

SDK for Python and JavaScript. MCP attachment for Claude and ChatGPT. A dedicated community node for n8n. Connect in minutes, log from the first decision. No schema changes, no new infrastructure, no compliance overhead pushed onto your sprint.

Compliance teams

Audit reports written in plain language with full regulatory citations. Know which framework applies, which signals passed, and which gaps need remediation, before the regulator asks, not after.

Procurement and risk

Your vendors are deploying AI agents that touch your data, your clients, and your decisions. Truveil produces the due diligence package you need to evaluate them, and the evidence trail your own clients will ask you to provide.

Enterprises

Dozens of agents. Multiple jurisdictions. One accountability layer. Truveil tracks every deployment, scores against the strictest applicable rule per market, and produces consolidated reporting your legal, risk, and board committees can act on.

See what a Truveil audit actually says.

Seven complete reports, published as they came out of the engine. Nothing was selected for its grade.

India · Email triage A 81.2
Email Triage Agent
Highest stake in the set: a real, irreversible action.
6 logged events · DPDP Rules 2025
India · Equity research C 62.8
MarketMind Equity Intelligence Agent
Multi-agent environment: three models, one accountability trail.
16 logged events · DPDP Rules 2025
European Union · Hiring B 68.6
Senior Data Engineer Hiring – Frankfurt
The only binding obligation in force today.
12 logged events · EU AI Act
See all seven reports

Advisory: regulatory expertise in two parts.

Truveil's advisory tools answer two distinct kinds of question. Both available free in Claude, ChatGPT, and any MCP-compatible client, including as an external tool in n8n agent nodes. Both grounded in primary regulatory text across six frameworks.

Regulatory advisory

Ask about laws, frameworks, jurisdictions. Get back clause-level analysis with primary-source citations.

Powered by consult_regulation and get_jurisdiction_brief.

Build advisory

Describe what you are building. Get back the regulatory obligations that apply, sequenced for action.

Built on Truveil Advisory, powered by Anthropic Claude.

"I'm building an AI loan approval system for the EU and India. What applies?"

Most consequential regulatory determination first. Provider/deployer call. Sequenced action plan. Cross-jurisdictional carry-forward.

"Does the EU AI Act apply to my US-based AI tool with EU customers?"

Clause-level analysis with primary citations.

"What's the DPDP audit timeline for a Significant Data Fiduciary in India?"

Rule 13 DPIA and audit cadence, SDF designation triggers, penalty exposure under DPDP Act Section 33.

"How does Singapore's Model AI Governance Framework treat generative AI?"

MGF-GenAI nine dimensions, AI Verify testing alignment, IMDA expectations for deployers and developers.

"Brief me on AI governance in the UAE."

Structured overview: Federal PDPL, DIFC Regulation 10, ADGM, UAE AI Charter. Effective dates. Binding instruments.

Get advisory access free

Advisory access is included in the free tier. After signup, you receive install instructions for Claude, ChatGPT, and any other MCP-compatible client.

Connect your agent in minutes.

Truveil integrates at the point of decision. Every log is structured, timestamped, and scored the moment it is written. Your audit trail starts on day one.

agent.py

        
Live feed

One decision, one record: subject, factors weighed, outcome, confidence. That is what a defensible audit trail is made of.

Why Truveil?

What makes Truveil different from adding a logging layer and calling it compliance.

Deterministic

The same evidence, scored against the same framework, produces the same result every time. No probabilistic drift, no hallucinated compliance claims. Truveil's output is reproducible and auditable by design.

Regulator-grounded

Scoring is grounded in primary regulatory text, not approximations. EU AI Act articles, NIST RMF functions, India DPDP obligations: each signal traces to a specific requirement in a specific instrument. Your audit report cites the source.

Model-independent

Truveil does not care what model your agent runs on. GPT-4o, Claude, Gemini, a fine-tuned open-source model: the accountability layer is the same. Switch models without losing your audit history or your compliance baseline.

Private Beta

Be first to hold your agent accountable.

Truveil is onboarding a small first cohort. Tell us what you are building and we will reach out with access.

Something went wrong on our side. Your details are still here; please try again in a moment.

Request received. Check your inbox for a confirmation, and we will follow up with access within 24 to 48 hours.

"Most people ask what AI can do. I spent my career asking how. That question built Truveil."

Karan Sharma, Founder
Read the full statement →

The audit trail you produce today is the one regulators read tomorrow.

Free tier includes 500 audit logs per month and unlimited regulatory advisory. Upgrade when you ship more agents.