AI Decision Engine — Live Data Pipeline

AI Decision Engine for Location-Independent Capital

Automated predictive execution, controlled by your risk limits.

Velnori analyses real-time market data and mirrors structured AI strategies directly to your brokerage. You set the boundaries. The system handles execution across time zones.

Configure Strategy Parameters Read methodology documentation →

Model Latency: 42ms
Ensemble inference

Risk Ceiling: 4.5% Max Drawdown
User-defined

Execution: Non-custodial API
Funds stay with broker

Mirrored Allocation — LiveActive Mirroring

Predictive Score0.74 Confidence

Risk BoundaryDrawdown 2.1% / 4.5%

Macro Volatility Arbitrage — 42%

Multi-Factor Equity — 35%

Liquidity & Yield — 23%

Rebalance Trigger: Volatility > 1.8σUTC Synced

Operational Context

Manual analysis does not keep pace with global markets.

When you manage capital across time zones, spreadsheets and delayed alerts create gaps. Velnori replaces fragmented checks with a continuous pipeline.

Manual constraints

Working asynchronously means decisions lag. Data stays scattered.

–Latency across zones. Signals arrive overnight. Action waits until morning.
–Fragmented sources. Prices, macro releases and sentiment sit in separate tools.
–Inconsistent risk checks. Stop levels depend on memory and manual review.

The Velnori autonomous pipeline

One system ingests, scores and executes within limits you define.

✓Real-time ingestion. Market feeds, macro indicators and sentiment processed continuously.
✓Predictive scoring. Ensemble models weight probability vectors before allocation.
✓Automated execution. Mirrored orders route via API only inside your risk ceiling.

Ingest → Score → Enforce → MirrorFour-stage control loop

42ms median inferenceMeasured at model gateway

24/5 market coverageAsynchronous by design

Technical Specification — STR-01 / STR-03

Strategy Catalogue

Three model classes. One mirroring engine.

Each strategy defines its inputs, risk ceiling and execution logic in advance. You select allocations and caps. The engine mirrors accordingly.

STR-01Status: Active Mirroring

Macro Volatility Arbitrage

Analyses volatility regimes and macro releases. Adjusts exposure when dispersion exceeds defined thresholds.

Model inputs: Vol indices, rates, CPI prints
Risk ceiling: 4.5% max drawdown
Execution logic: Reduce exposure if volatility > 1.8σ

STR-02Status: Active Mirroring

Multi-Factor Equity Optimisation

Scores equity baskets on value, momentum and quality factors. Rebalances toward higher probability-weighted mixes.

Model inputs: Fundamentals, price momentum
Risk ceiling: 3.0% max drawdown, sector caps
Execution logic: Weekly rebalance, drift > 2%

STR-03Status: Active Mirroring

Liquidity & Yield Rebalancing

Monitors cash, money-market and short-duration yield. Keeps liquidity buffers intact while optimising idle capital.

Model inputs: Yield curves, FX, spreads
Risk ceiling: 1.5% max drawdown, £ cap
Execution logic: Daily sweep, minimum £500 tranche

Model outputs are simulations based on defined parameters and live data. They are not guarantees of performance. All mirroring respects hard stop-loss thresholds and exposure caps. Capital remains subject to market risk.

Decision Architecture

How a signal becomes a controlled order.

Four auditable stages. Each stage logs inputs and decisions for review and export.

01 — INGEST

Multi-source data ingestion

Market feeds, macro indicators and sentiment streams are normalised and timestamped in UTC.

Feeds: equities · FX · rates

02 — INFER

Predictive inference

Ensemble models score probability vectors and rank allocations by expected risk-adjusted behaviour.

Latency: 42ms median

03 — ENFORCE

Risk boundary enforcement

Drawdown limits and exposure caps are checked before any order. Breaches block execution.

Ceiling: 4.5% max drawdown

04 — MIRROR

Automated strategy mirroring

Approved allocations route via broker API and rebalance automatically. You can pause at any time.

API: non-custodial execution

Risk Governance

You retain full control over capital safety.

Velnori is non-custodial. Funds stay with your broker. The platform only sends orders inside guardrails you configure. When limits are hit, mirroring pauses.

Set conservative thresholds when travelling. Review logs asynchronously. Adjust caps without closing positions. The kill-switch halts all new orders immediately while leaving existing holdings untouched for manual review.

Configurable parametersApplied pre-execution

Parameter	What it controls	Example setting
Max drawdown	Hard stop for each strategy	4.5% then pause
Capital allocation cap	Maximum £ per model	£10,000 per strategy
Asset exclusions	Instruments to never hold	Exclude leveraged ETFs
Kill-switch trigger	Immediate halt condition	Volatility > 2.5σ

Velnori risk governance dashboard for remote portfolio control

Guardrails are evaluated before every order. Logs show which rule approved or blocked each action.

Explore Predictive Models

Decision Support

Practical answers for UK-based and mobile operators.

Short explanations of custody, execution, risk and reporting.

How is my capital held? Does Velnori take custody?

No. Velnori is non-custodial and connects via read-and-trade API keys. Funds remain in your brokerage account. You can revoke access at any time. Withdrawals are only possible from your broker.

How does AI strategy mirroring execute without manual input?

Once you configure allocations and caps, the engine scores models continuously. When a rebalance condition is met and risk checks pass, orders route automatically via API. Every decision is logged with timestamp, model version and rule outcome.

Can I customise risk thresholds for volatile markets?

Yes. Set max drawdown, per-strategy £ caps, asset exclusions and kill-switch triggers. Tighter settings reduce activity. Breaches pause mirroring rather than closing positions automatically, unless you enable auto-flatten.

What data sources are used, and how often are models retrained?

Models analyse exchange feeds, macro releases, yield and FX data, plus aggregated sentiment. Features update intraday. Model weights are reviewed on a fixed cycle and versioned, so you can see which version drove each allocation.

Can I export data for UK tax and accounting?

Yes. Export trade logs, holdings, fees and realised P&L in CSV. Reports use GBP where applicable and include UTC timestamps to support Self Assessment records and accountant review. Velnori does not provide tax advice.

Onboarding

Deploy automated data models in three steps.

Connect your broker via secure API. Configure caps and exclusions. Activate mirroring and monitor logs from any time zone.

Start Configuration Compare operating models

No custody transfer. Pause or revoke access at any point. UK support hours with async log review.

1Connect.
Link brokerage with scoped API keys. Funds stay in place.

2Configure.
Set drawdown caps, £ limits and excluded assets.

3Mirror.
Activate strategies. Orders execute inside your boundaries.