Agents gather
Filings, transcripts, news, market structure, and event odds — read in full, around the clock, across your entire universe.
Scalar Field for Hedge Funds
Scalar Field is the agentic AI platform for the front office. Market data, filings, news flow, and prediction markets unified in one command center — where always-on agents research, monitor, and act on your process, not a generic one.
CombinatorMost investment teams spend 80% of their time gathering information and 20% acting on it. Agents invert that ratio: they do the gathering continuously, the synthesis instantly, and leave the judgment — the part that actually compounds — to you.
Filings, transcripts, news, market structure, and event odds — read in full, around the clock, across your entire universe.
Analysis with every claim traceable to its source. No black-box summaries, no hallucinated numbers.
Your time goes to sizing, timing, and conviction — the decisions that actually pay for the seat.
Everything an investment team expects from an institutional AI platform — delivered as workflows, not features.
Ask a complex question, get institutional-grade analysis in minutes. Agents write and execute code against point-in-time data — every number traceable to its source.
Screen the entire market on fundamentals, positioning, short interest, insider activity, and sentiment — in plain English, as often as your process demands.
SEC filings, earnings transcripts, and presentations parsed the moment they drop. Extract what moves your thesis and skip the rest.
Always-on agents watch news flow, filings, and market structure for the setups you care about — so nothing slips past your coverage.
Stress-test a book against rate paths, macro shocks, and event outcomes. Model the scenario with agents before the market forces you to.
Exposure, risk, and performance attribution — answered conversationally, computed on live positions instead of last night’s export.
News Agents
Always-on news agents read the entire flow, every hour of every session. You define what matters in plain English, and the moment a trigger fires you get an instant notification — not a digest the next morning.
“Notify me if an activist files a 13D on any name in my book.”
“Alert me the moment a company in my coverage — or one of its suppliers — cuts guidance.”
“Ping me when tariff or export-control headlines touch my semiconductor supply chain.”
“Tell me if Polymarket odds on a December cut move more than 10 points in a day.”
Triggers are yours to define — any name, any theme, any relationship in your book. Agents evaluate every story against them and stay silent until one fires.
Prediction Markets
Elections, Fed decisions, geopolitics — prediction markets price events before sell-side notes do. Scalar Field is the only platform in this space where agents can query, analyze, and act on them, wallet by wallet.
Ask anything about Polymarket and Kalshi markets — odds history, volume, liquidity, and how probabilities are moving right now.
Follow wallets with proven win rates on-chain. Get notified the moment they place high-conviction bets, and see their full position history.
Compare event probabilities against how equities, options, and crypto are pricing the same event — and find the gap before consensus does.
Take the trade live or on paper from the same platform — prediction markets are an execution venue here, not just a data row.
The Agent Harness
The news agents and prediction market agents above run on the Scalar Field agent harness — the same infrastructure we use to encode your firm's own workflows into custom agents.
Agents are encoded with how your firm models businesses and where you seek edge — your checklists, your frameworks, your definition of a setup. Not a generic AI wrapper.
Run any frontier LLM — OpenAI, Anthropic, Google, or your own endpoints — and swap models on the fly. The harness is yours; the model is a choice.
Agents run on schedules and monitor continuously. Every name in your universe gets watched every day — not just the ones an analyst had time for.
Institutional AI lives or dies on data handling. Scalar Field is built so the answer to every due-diligence question is the one you want to hear.
Independently audited security controls, built to meet institutional due-diligence and DDQ requirements.
Single-tenant deployment available. Your research, positions, and queries never leave your environment.
Zero-retention agreements with LLM providers. Prompts and outputs are never stored by the model vendor.
Your data never trains any model — ours or anyone else’s. Full stop.
Encryption in transit and at rest, with complete isolation between clients.
Every insight is traceable to source data. When data is unavailable or uncertain, agents say so instead of guessing.
Institutional-grade market data unified in one place — including the sources conventional stacks don't carry, like prediction markets and on-chain wallet flows. No separate data contracts required to get started.
For investment teams
The full agentic workspace: deep research, screening, document intelligence, news triggers with instant notifications, and prediction market intelligence — ready for your analysts on day one.
For quant and technology teams
The same data and agent infrastructure, integrated into your own stack — REST API and Python library for research pipelines, internal tools, and systematic workflows.
Every enterprise engagement includes white-glove onboarding: we encode your workflows into agents alongside your team, and we stay until your analysts are measurably faster.
Scalar Field is available as a managed cloud platform or as a private, single-tenant deployment for firms that require full data isolation. In a private deployment, your research and queries never leave your environment.
Nothing leaves your control. Scalar Field is SOC 2 audited, maintains zero-data-retention agreements with LLM providers, never trains models on client data, and encrypts all data in transit and at rest with complete isolation between clients.
Scalar Field is model-agnostic. Agents can run on any frontier LLM — OpenAI, Anthropic, Google, or your own private endpoints — and models can be swapped without changing your workflows, so there is no vendor lock-in.
You define triggers in plain English — for example, "notify me if an activist files on any name in my book." Always-on news agents read the full news flow, evaluate every story against your triggers, and send an instant notification the moment one fires.
Agents can query and analyze Polymarket and Kalshi markets, track smart-money wallets with proven win rates, compare event probabilities against how assets price the same event, and act on the signal live or on paper.
Custom agents are built on the Scalar Field agent harness and encoded with how your firm models businesses and seeks edge. Enterprise onboarding is white-glove: we build your first agents alongside your team.
Book a 30-minute demo. Enterprise engagements typically begin with a pilot in which we encode two or three of your workflows into agents, so your team evaluates Scalar Field on your own process rather than a canned demo.
Book a 30-minute demo. Bring a workflow your team runs every week — we'll show you what it looks like when an agent runs it every hour.