Introducing

ORION

Agentic research for systematic and discretionary investors — from literature and screens to backtested signals and investment counsel, unified by a shared Idea Vault. Built on your infrastructure with governed agent pipelines.

Waitlist open

Join the waitlist See how it works →

Designed by institutional systematic equity PM — the research flow reflects how production quant teams actually operate. Preview workspace mockup →

Product demo

See ORION in action

Live research workspace walkthrough — literature search, paper analysis, knowledge graph, and hypothesis planning. (~3.5 min)

Today’s systematic research workflow involves a lot of manual stitching: idea generation is ad hoc or spreadsheet-based, backtests run in disconnected systems, and signal delivery requires someone to package and route output to the right desk. ORION replaces the stitching with a single governed pipeline — from hypothesis kickoff to portfolio-ready signal — without requiring you to migrate your data or change your execution layer.

IDEA GENERATION

Hypothesis Engine

Most idea generation is either discretionary (subject to recency bias) or limited to the data sources you’ve already integrated. ORION’s hypothesis engine synthesizes alternative data, macro signals, cross-asset flows, and sentiment continuously — agent-curated and ranked by signal strength and historical reliability, so the queue reflects what the data suggests, not what you’ve been watching.

SURVIVING BACK TESTING

Systematic Strategy Evaluation

Manual backtesting is slow, parameter-sensitive, and biased toward the regimes you thought to test. ORION runs automated evaluation across historical regimes with agent-driven parameter search — outputs ranked by Sharpe, max drawdown, and regime-conditional performance. You see the strategies that survive the regimes you didn’t anticipate.

SIGNAL DELIVERY

Portfolio-Ready Output

Signal packaging and routing is typically manual — someone exports, formats, and routes output to the right desk or system. ORION delivers signals directly to your execution layer or reporting stack, adapting to your existing infrastructure. No new systems. No forced migration. Signals arrive in the format your team already works with.

ORION is built for

Systematic equity and quant macro teams who run research across disconnected tools
Portfolio managers building or formalizing a systematic process who need a governed pipeline
Research teams at regulated firms that need auditability baked into the research flow

ORION is not (currently) for

Discretionary managers with no interest in systematic signal generation
Teams looking for a fully managed, cloud-hosted SaaS with no infrastructure touchpoints
High-frequency or ultra-low-latency execution use cases

Workflow preview

Idea generation, surviving backtesting, and signal delivery — three governed stages in one pipeline.

Idea Generation Part 1 — Synthesize Ideas (~4 seconds at 0.75×) The hypothesis engine synthesizes alternative data, macro signals, and cross-asset flows — agent-curated and ranked so the idea queue reflects what the data suggests, not recency bias.

Surviving Back Testing Part 2 — Backtest analysis (~34 seconds at 1.5×) Automated backtest runs across regimes — trade distribution, drawdown profile, and agent critique of strategy design. The pipeline surfaces what survives statistical scrutiny before anything advances.

Signal Delivery Part 3 — Publish Signal to IC and staging bootstrap for production (~4 seconds at 0.75×) Ranked signals publish to the investment committee workflow, then bootstrap into staging for production routing — Sharpe-sorted, regime-labelled, with a human review checkpoint before anything leaves the pipeline.

The platform is live — explore Analysis, supply chain intelligence, and our research libraries.

Explore today or book a consultation

See live demos today → View pricing →