Every AI tool call — checked, controlled, and audited before it executes.
ConsentGuard wraps any AI agent in RBAC, risk policy, and immutable audit trails. One line of code. 60-second integration. Built for fintechs, NBFCs, and banks.
Or join the early access waitlist
Agent requested write access to Stripe API
Execution control for every agent action
Deterministic policy gates run before tool calls touch production systems.
Stop unauthorized actions before they run.
RBAC, risk scoring, and rule templates run on every tool invocation. A compromised agent attempting a bulk export, an unauthorized transfer, or a destructive delete is blocked outright - not logged after the fact.
Route high-impact actions to humans.
Drop-in React consent modal, plus email and Slack approval flows. Every consent decision carries structured rationale: who approved, why, and under which policy. Replay-ready for audit.
Audit trails mapped to RBI FREE-AI and DPDP.
Every decision - allow, block, or consent - is persisted with structured reasons, trace IDs, and context snapshots. Export-ready for regulators and internal risk teams.
Three forces are converging
RBI FREE-AI Framework
India's Reserve Bank now requires human-in-the-loop controls for AI agents handling financial transactions.
DPDP Act 2023
Explicit consent and auditable decision trails are no longer optional for production AI systems.
AI Agents Hit Production
Enterprise AI agent spend is accelerating, and governed execution is becoming the limiting factor.
A governance layer operators can understand
Real-time approvals, clear escalation paths, and searchable audit trails instead of opaque agent behavior.
Governed execution timeline
Searchable audit evidence
Every approval, denial, policy decision, and execution result is stored in a format compliance teams can review fast.
Policy-aware approvals
Thresholds, actor roles, and allowed tools are enforced consistently instead of hand-coded into every workflow.
Three UI surfaces teams can embed fast
ConsentGuard ships with opinionated UI primitives for consent, approvals, and governed agent execution across lightweight and full-session workflows.
Contextual Pill
Non-intrusive safety indicator.
Floating Chatbot
Drop-in chat with built-in consent controls.
I need approval to process a refund of $49.00 for customer@example...
Sidesheet Assistant
Deeply integrated panel for complex data analysis tasks and multi-turn loops.
This action requires elevated privileges. Do you want to proceed?
Interactive demos and component surfaces
Jump directly into the governed assistant flows and the higher-risk fraud response scenario.
Governed Copilot Demo
Simulate low-risk reads, approval-gated transfers, destructive actions, and policy-blocked requests in a live assistant flow.
Approval modal, audit trail, and governed tool execution states.
Shows the core SDK interaction model and operator-facing UI surface.
High-Risk Response Demo
Walk through escalation, denial, account freezes, and SAR-report generation in a multi-step fraud containment sequence.
Critical approvals, incident-response logs, and multi-step state transitions.
Shows the more advanced operator UI for regulated, high-severity workflows.
Zero-Config Integration
Get up and running in minutes, not months.
Initialize the Client
Import the SDK and wrap your LLM provider in a single line of code.
import { ConsentGuardClient } from 'consentguard';
const client = new ConsentGuardClient({
apiKey: process.env.CONSENTGUARD_API_KEY
});Define Governance Policies
Set RBAC, risk, and consent rules for governed execution.
const safeTool = wrap(sendEmail, {
client,
riskLevel: 'high',
consentFlow: 'email',
auditTag: 'customer_email'
});Deploy with Confidence
Launch AI features knowing ConsentGuard is auditing every step.
const res = await agent.chat(input);
// Automatically governed & loggedReady to scale?
Transparent pricing for teams of all sizes.
- 10,000 tool calls/mo
- RBAC + consent flows
- RBI FREE-AI templates
- Audit log export (CSV)
- Email support
- Unlimited tool calls
- Advanced risk rules engine
- Custom compliance templates
- Audit export (PDF + CSV)
- SSO + team management
- 24/7 Slack support
- Unlimited everything
- Data residency options
- Custom SLAs
- Dedicated compliance engineer
Technical FAQ
Syed Ajwad
Founder, ConsentGuard
Why I'm Building This
Ready to govern your AI?
Be among the first to govern your AI agents.
Join the early access waitlist