Architecture Blueprint

Live Risk Architecture Blueprint

Optional patterns for production live risk systems. Use what works for you.

Looking for Basic Integration?

This page covers advanced architecture patterns. If you're just getting started, see our simpler Integration Guide.

AADC Integration Guide

This Page Is For You If...

You've already integrated AADC and now you're asking:

  • ? How should we structure kernels for a full trading desk?
  • ? What's the best way to handle portfolio updates during the day?
  • ? How do we manage kernel lifecycle in production?
  • ? What architecture enables sub-second live risk?

This is NOT a prerequisite

Many customers run AADC successfully without any of this. But if you're building a production live risk system, these patterns will save you months of trial and error.

A Note on Complexity

You don't need to 'divide your library into kernels'

AADC records whatever function you execute. If you call `price_portfolio()`, that becomes a kernel. If you call `calibrate_curves()`, that's another kernel.

The patterns here are descriptive, not prescriptive

This blueprint answers: 'What functions do successful live risk implementations typically record?' It's not a requirement — it's a collection of what works.

Start simple

  1. 1 Record your main pricing function
  2. 2 See the speedup
  3. 3 Add more kernels as needed

Most customers start with a single kernel and expand organically. The architecture emerges from your actual workflows, not from upfront planning.

Common Kernel Organization Patterns

Here's what successful implementations typically look like — not what you must do

AADC records whatever function you execute—no upfront planning required. As your usage grows, you'll naturally develop patterns that work for your workflow. Many teams end up with separate kernels for curve calibration, portfolio pricing, and risk sensitivities, but this emerges organically from actual use.

Model Calibration Kernels

Most teams record curve-building as a separate kernel. Why? Because curves change less frequently than portfolios.

Typical: Generated once per day or on significant market moves

  • • IR curve bootstrapping
  • • Vol surface calibration
  • • Credit curve generation

Portfolio Pricing Kernels

Record your portfolio pricing once, execute thousands of times with different market data throughout the day.

Typical: Recorded at start of day, reused for every price update

  • • Full portfolio valuation
  • • Desk-level aggregations
  • • Product-specific pricing

Risk Sensitivity Kernels

Greeks come automatically via the adjoint pass. Some teams combine pricing and risk; others keep them separate.

Typical: Often integrated with pricing, sometimes separate for flexibility

  • • Delta/Gamma/Vega
  • • Cross-gamma matrices
  • • Scenario sensitivities

Daily Trading Lifecycle

A typical day with AADC kernels

Here's how production teams typically structure their day. This isn't a requirement — just a common pattern.

Pre-Market (5:00-7:00)

Curve Calibration

Generate calibration kernels from overnight data. These run fast enough that you can recalibrate whenever curves move significantly.

Market Open (7:00-8:00)

Portfolio Recording

Record portfolio pricing kernels for the day's positions. New trades added during the day can extend existing kernels.

Intraday (8:00-17:00)

Continuous Execution

Execute pre-recorded kernels with live market data. Sub-millisecond updates. No re-recording needed for market moves.

Throughout Day

RFQ Handling

Standard trades use pre-recorded kernels (sub-millisecond). Bespoke trades get recorded on-the-fly (seconds, then cached).

Intraday Pricing workflow

Trade Amendments Without Re-Recording

One of AADC's key advantages

You don't need to regenerate kernels when trades change. Use scaling factors to add, remove, or modify trades in place.

Adding Trades

Append new trades to existing kernels. The kernel structure accommodates new positions without full regeneration.

How: New trade inherits kernel structure; scaling factor starts at 1.0

Removing Trades

Set a trade's scaling factor to zero. The kernel still runs; that trade simply contributes nothing.

How: Scaling factor → 0 removes contribution; kernel remains valid

Partial Unwinds

Reduce position size by adjusting the scaling factor. Half the position? Set scaling to 0.5.

How: Scaling factor 0-1 for proportional position changes

Parameter Changes

Many parameter changes don't require re-recording at all. Market inputs can be updated directly.

How: Input parameters can be changed without touching the kernel

Trade Amendments workflow

What-If and Scenario Analysis

Same kernels, different inputs

Once you have a kernel, scenario analysis is essentially free. Apply different market conditions to see the impact.

Market Scenarios

Run the same kernel with stressed market inputs. What happens if rates move 100bp? Same kernel, different inputs, instant answer.

Trade What-Ifs

Combine scaling factors with scenarios. 'What if we added this trade AND rates moved 50bp?' Handled in milliseconds.

Stress Testing

Run regulatory stress scenarios across the entire portfolio. The kernel handles thousands of scenarios without re-recording.

Deployment Options

Binary kernels run anywhere

AADC compiles your models to binary kernels — optimized machine code that runs 6-1000x faster. These kernels are portable.

On-Premises

Run entirely within your infrastructure. Full control, no external dependencies.

  • Complete data sovereignty
  • No external connectivity required
  • Integrate with existing grids (LSF, SLURM)

Best for: Regulated environments, latency-sensitive applications

Cloud Execution

Deploy encrypted kernels to AWS, Azure, GCP. Your models stay on-premises; only compiled binaries move to cloud.

  • Up to 99% cost reduction vs traditional cloud
  • Burst capacity on demand
  • Source code not included in kernels

Best for: Scalability, cost optimization, burst capacity

Hybrid Architecture

Record on-premises where your IP lives. Execute in cloud when you need scale. Best of both worlds.

  • IP stays on-premises
  • Cloud scalability when needed
  • Automatic kernel transfer

Best for: Most production deployments

The Key Insight

Record once on-premises where your proprietary models live. Deploy kernels anywhere — cloud, edge, multi-region. Your IP never leaves your secure environment.

Deployment Architecture

What Live Risk Architecture Enables

Ultra-Low Latency

Sub-millisecond pricing for standard trades

Efficient Amendments

Trade changes without kernel rebuilds

Complete Greeks

All sensitivities in single pass

Scalable

From single desk to enterprise-wide

Related Resources

Frequently Asked Questions

Do I need to plan kernel architecture before using AADC? +
No. AADC records whatever function you execute — no upfront planning required. Most customers start with a single kernel recording their main pricing function and expand organically based on actual needs. The patterns in this blueprint describe what successful implementations look like; they're not prerequisites.
How do I handle trade amendments without regenerating kernels? +
AADC kernels support scaling factors for each trade. To remove a trade, set its scaling factor to zero. To add a trade, append it to the kernel with scaling factor 1.0. For partial unwinds, use fractional scaling factors. This avoids costly kernel regeneration for routine trade changes.
What's the difference between this page and the Integration Guide? +
The Integration Guide covers the basics: how to get AADC working with your code in weeks. This page covers optional advanced patterns for teams building full production live risk systems. Most users should start with the Integration Guide; this page is for when you're ready to scale.
Can I deploy AADC to the cloud while keeping my models secure? +
Yes. With hybrid deployment, you record kernels on-premises where your source code lives. Only binary kernels move to the cloud — source code is not included. Your source code stays secure while you get cloud scalability.

Ready to Design Your Live Risk System?

We can help you plan the architecture that fits your specific needs — or just answer questions about what's worked for others.