Simudyne

Horizon

Agentic workflows for financial services

Automating financial R&D for risk modelling, portfolio analysis, and equity valuation.

Products

Two modules, one platform

Simudyne Horizon combines advances in agentic frameworks with recent research in quantitative finance and AI.

Horizon :: Risk

Early access Q1 2026

Agentic discovery of risk factors and automated portfolio analysis. Horizon Risk builds bespoke, dynamic risk models for your portfolio, continuously monitoring and proposing new factors as market conditions change.

  • Automated risk factor discovery using agentic AI
  • Bespoke portfolio-specific risk models
  • Sparse factor graphs for systemic risk analysis
  • Dynamic model updates with daily monitoring
  • Returns attribution and performance reporting

Horizon :: Value

Coming soon

Agentic valuation reports for equity markets. Horizon Value automates fundamental analysis workflows, generating research-grade valuation reports and identifying mispricings across sectors and geographies.

  • Automated equity valuation reports
  • Fundamental analysis workflows
  • Sector and regional screening
  • Scenario analysis and stress testing
  • Integration with existing research processes

Capabilities

Innovations in agentic and risk modelling

Agentic Discovery Framework

The Horizon discovery loop automates the scientific process of model building into finance through an agentic version of the builder-critic loop. Generative AI agents iterate on model construction, evaluation, and refinement without manual intervention.

Dynamic Risk Factor Models

Risk models are bespoke to a specific portfolio and dynamic, updating to market changes and giving portfolio-specific insights. Standard methodology is augmented with statistical risk factors discovered by the agentic framework.

Sparse Factor Graphs

Horizon outputs a common factor model, a structured latent factor model, and a sparse conditional dependency graph. This graph encodes which assets share direct risk linkages beyond common factor exposures, revealing hidden systemic connections.

Automated Risk Research

Daily reports flag changes in risk model explanatory power and propose new risk factors. Users review proposed factors, compare against current models, and decide whether to integrate them into their portfolio monitoring.

Developer first

Risk factors in five lines of code

Generate, evaluate, and integrate risk factors into your risk reports programmatically. The Horizon SDK handles discovery, model construction, and reporting.

example.py
from simudyne import Horizon

client = Horizon(api_key="...")
portfolio = client.risk.analyze("portfolio.csv")
factors = client.risk.discover_factors(portfolio)
report = client.risk.generate_report(portfolio, factors)
print(report.summary)

Early access

Join the waitlist

Horizon :: Risk is available for early testing. Leave your details and we will be in touch.

FAQs

Frequently asked questions