On the tape
▶
On the tape
On the tape
Your Agentic Trading Desk - Building the next era of agentic…
Company intel
We’re building the agentic trading desk for the new operating system of markets. I used to be a trader at Tower Research and Goldman Sachs, and I’m building Scalar Field with my co-founder Ramakant, who led engineering teams at Microsoft. We’ve both spent our careers at the intersection of finance and technology, and we’re using that experience to rethink how markets are analyzed, tested, and traded in the age of agents. Scalar Field is built around a simple belief: over the next few years, agents will not just help people research markets - they will actively participate in them. They will monitor news, parse filings, track sentiment, test hypotheses, rebalance portfolios, and execute trades. They will turn ideas into strategies, and strategies into live portfolios. But today’s financial infrastructure was not built for that world. Terminals are dashboards. Backtesting tools are fragmented. Execution systems are separate. Market data is hard to unify. And most platforms still assume the human is the only decision-maker. Scalar Field changes that. We’re building an agentic trading platform where users can create financial agents that research data, backtest strategies, monitor market events, and trigger trades when specific conditions are met. The long-term vision is agentic ETFs: portfolios created, managed, and rebalanced by agents around any idea, narrative, event, or market signal. Scalar Field lets users go from thesis to backtest to live portfolio in one place. It is infrastructure for the agentic trading era.
We’re launching Scalar Field - the agentic trading desk for the new operating system of markets. Scalar Field lets you go from an investment idea to research, backtesting, and live execution in one place. We believe trading agents will do much more than answer questions about markets. They will monitor data, analyze events, test strategies, manage portfolios, and execute trades across multiple markets. But trading agents cannot be built like standard AI agents. Most agent frameworks place an LLM call in the critical path of every action. That approach is too costly, too slow, and too nondeterministic for many trading workflows. Scalar Field separates reasoning from execution. Agents can use LLMs to understand filings, news, market narratives, and unstructured information, while deterministic and model-driven systems continuously monitor signals and execute trades without requiring an LLM call every time. Our agents can achieve event-to-trade latencies of approximately 300 milliseconds. Scalar Field currently supports: • Robinhood for US equities and ETFs • Public.com for US equities and options • Polymarket for prediction markets • Jupiter DEX for Solana assets and tokenized equities • Hyperliquid support is coming soon • Alpaca for US equities and options You can build agents that: • Track which companies are increasing AI capital expenditure and continuously rebalance a portfolio around them. • Monitor insider purchases, test whether trades by CEOs or directors predict future returns, and execute the resulting signal. • React to earnings, inflation releases, interest-rate decisions, breaking news, or changes in prediction-market probabilities. • Monitor live sports games and trade when the market overreacts - for example, when the starting favourite falls behind but its win probability begins to recover. • Run simple deterministic strategies without using an LLM at all. • Train transformer models on historical market data, generate signals in real time, and automatically execute trades based on those signals. Our long-term vision is agentic ETFs: portfolios created, managed, and continuously rebalanced by agents around any idea, narrative, event, or market signal. Trading agents need more than a chatbot connected to a brokerage API. They need financial data, research compute, persistent state, portfolio infrastructure, risk controls, and low-latency execution. That is what we’re building at Scalar Field. Scalar Field: YC Spring 2025: Scalar Field | The agentic trading desk
Sector
Tags
Led by Y Combinator
Est. 2026 · Study the cut
