For professional crypto desks
Controlled agentic crypto trading for professional desks
Deploy AI-assisted trading workflows with clear mandates, enforced risk limits, approval controls, multi-venue routing, and audit evidence. Rugira governs what agents may do before any action reaches an exchange.
CONTROLLED AGENTIC WORKFLOW
How controlled agents work
Agents may observe markets, portfolios, and venues. Rugira converts their proposals into structured trade intents, checks them against your mandate, escalates to human approval when required, routes only approved orders, and preserves every step for audit.
Observe
Agents monitor markets, portfolio state, venue health, and policy context.
Propose
A trade idea becomes a structured proposal with size, venue, rationale, and confidence.
Check
Rugira validates the proposal against mandate, risk limits, kill-switch rules, and venue constraints.
Approve
Human review is required when your desk policy demands maker-checker approval.
Execute
Approved orders route through trade-only API keys and existing venue connections.
Audit
Proposal payloads, policy checks, approvals, routing details, and fills are preserved for review.
Who it's for
Who it's for?
Funds / Quants
Small funds and quant teams can turn trading logic into supervised agent workflows. Simulate, approve, and deploy strategies with hard policy checks before execution.
Agent mandates convert strategy logic into enforceable operating boundaries.
Controlled agentic workflow
See a trading agent act under control
Proposal to execution, under policy
A short walkthrough of how a supervised trading workflow moves from market observation to approval, routing, and audit evidence.
Watch an AI-assisted trading workflow identify a market opportunity, generate a structured proposal, pass through Rugira’s policy gate, request approval when required, route execution, and produce a complete decision trail.
Agent mandate
4 controls activeAgent Mandates
Define which venues, symbols, sizes, time windows, and conditions each supervised trading agent may act on.
Define which venues, symbols, sizes, time windows, and conditions each supervised trading agent may act on.
Policy-Gated Execution
Every proposed action is checked against limits, readiness checks, venue constraints, kill-switch rules, and approval requirements.
Every proposed action is checked against exposure limits, readiness checks, venue constraints, kill-switch rules, and approval requirements before execution.
Policy gate - active
Decision trail export
Agent Decision Trails
Capture proposal payloads, policy checks, approval actions, routing details, rejections, fills, and manual overrides.
Capture proposal payloads, policy checks, approval actions, routing details, rejections, fills, and manual overrides for audit and review.
Controls & governance
Security and control for supervised agents
Rugira is built for governed execution. Client assets remain with client-selected exchanges or custodians. Agents operate through trade-only credentials, tenant-scoped isolation, MFA/SSO, role controls, approval workflows, and exportable audit evidence.
Non-custodial by design
No. Rugira is non-custodial. Orders execute through your existing exchange or custody accounts using trade-only credentials and your configured policies.
Your keys, your assets - Rugira never touches withdrawals
Non-custodial by design
Rugira does not hold client assets. Execution runs through your existing accounts and trade-only API keys.
Your policies control the agent
Define what each agent can observe, propose, and execute. Rugira blocks actions outside the approved mandate.
Frequently Asked Questions
No. Rugira is non-custodial. Orders execute through your existing exchange or custody accounts using trade-only credentials and your configured policies.
No. Rugira provides software controls, execution infrastructure, and reporting. Your desk defines strategies, mandates, limits, and approval policies.
Only within the mandate, policy, and approval framework configured by your desk. Live autonomy is optional, bounded, and auditable.
Rugira records proposal payloads, policy checks, approvals, routing details, execution outcomes, rejections, and audit exports.
Controlled agentic trading means AI-assisted workflows can monitor markets and generate trade proposals, but execution is governed by defined mandates, risk limits, approvals, and audit trails.
No. Rugira early access is reviewed manually and prioritized for professional desks with clear controlled-execution use cases.
Rugira is built for professional workflows. These details help us prioritize relevant use cases and avoid retail or unsupported requests.
Regulatory context
What regulators will expect from agentic trading
As AI-assisted execution becomes more common, professional desks will need to prove which policy applied, who approved the action, what the system observed, and why the final order was allowed or rejected.
The issue is not whether AI can propose a trade. The issue is whether your desk can prove how that proposal was controlled.
"In some markets, algorithmic trading already accounts for around 70% of total orders."
"Transaction monitoring is crucial for the prevention of financial crime, and the FCA expects firms to implement robust systems and controls."
"Income from digital assets is taxable, and you may need to report your transactions on your tax return."
Pricing
Plans & Pricing (CHF)
Controlled agentic trading plans for professional desks that need mandates, approvals, and audit evidence.
Agent Desk
For solo PMs and very small desks.
Annual prepay: 5'085–
- 2 supervised trading agents
- Paper and live modes
- 2 connected venues
- Agent mandate templates
- Basic simulation / backtesting
- Policy-gated execution
- Audit export
- Standard support
Agentic Ops
For operating desks with formal controls.
Annual prepay: 10'185–
- Up to 10 supervised agents
- Up to 5 connected venues
- Staging environment
- Maker-checker approval workflows
- Advanced simulation / backtesting
- SSO / team roles
- Advanced decision trails
- Priority support
Private Agent Gateway
For funds, brokers, and fintech platforms.
One-time setup: 5'000–20'000–
- Private deployment / VPC option
- White-label agentic trading workflows
- Custom policy modules
- Advanced approval chains
- Broker integration support
- External agent gateway readiness
- Custom compliance and reporting pack
- Dedicated onboarding
Annual prepay and setup fees remain available for professional and enterprise plans. VAT not included.
Rugira provides non-custodial software controls, execution infrastructure, and reporting. No investment advice.
Ready to control AI-assisted trading workflows?
See how Rugira lets your desk deploy supervised trading agents across existing exchange accounts - with policy limits, approvals, and audit evidence enforced by default.