Engineered for precision allocation
Every module of Bivori Noloru is built around one objective: translating raw market signal into disciplined, repeatable portfolio decisions — without management or performance fees eating into the result.
Four systems, one allocation engine
Each module runs independently but shares a common data layer, so signal from one feeds directly into the decisions of the others.
Neural Risk Modelling
Continuously re-scores exposure across the book using pattern recognition trained on historical volatility clusters, not static risk brackets.
Liquidity Analysis
Order book depth and spread behaviour are checked before any rebalance is executed, reducing slippage on size.
Adaptive Allocation
Position weights shift along a defined glide path as conditions change, rather than staying fixed to a quarterly schedule.
Zero-Fee Structure
No management fee and no performance fee layered on top of returns — the engine's output is the net result you see.
From signal to rebalance in four steps
-
Ingest & normalise
Market, liquidity, and volatility data are pulled from multiple venues and normalised into a single feed.
-
Score risk
The neural risk layer assigns a current exposure score to every held and candidate position.
-
Check liquidity
Proposed trades are filtered against live depth and spread thresholds before being queued.
-
Rebalance
Approved adjustments are executed and logged, updating the portfolio's allocation record.
Engine Parameters
Parameters reflect the general operating configuration of the current engine version and are subject to change as models are retrained.
What each module actually does
A closer look at the inputs each system relies on and the output it produces for the allocation engine.
| Module | Primary Input | Output | Refresh Behaviour |
|---|---|---|---|
| Neural Risk Modelling | Volatility & correlation data | Exposure score | Continuous |
| Liquidity Analysis | Order book depth, spreads | Trade feasibility | Pre-trade check |
| Adaptive Allocation | Risk scores + glide path | Target weights | Threshold-triggered |
| Reporting Layer | Execution logs | Allocation record | Per rebalance event |
Figures and behaviours described are general and illustrative of current engine design; they are not a guarantee of future performance or availability.
Designed around capital preservation first
Allocation shifts are gated by liquidity checks before anything is executed, which is deliberate: the engine is built to avoid forcing trades into thin markets, even when a risk score calls for an adjustment.
No fee layer sits between the engine's decisions and the account balance — the structure was kept flat so performance isn't diluted by recurring charges.
What this means in practice
Translating the technical modules into practical outcomes for anyone running a portfolio through Bivori Noloru.
Fewer blind spots
- Exposure is re-scored continuously, not on a fixed calendar
- Correlation shifts are picked up as they develop
- No reliance on a single static risk bracket
Cleaner execution
- Trades are checked against live depth before queuing
- Spread conditions are factored into timing
- Designed to reduce avoidable slippage on size
No fee drag
- Zero management fee on the account
- Zero performance fee on gains
- Net output reflects the engine's actual work
See the engine running on your terms
Get access to the terminal and review how the allocation logic applies to a live book.