Ithaca

Company: Ithaca

Self-reported basis: The entire analysis is based on the author’s own description of Ithaca as submitted to the OpenAI 2026 hackathon on Devpost. No independent verification or third-party data is available.

What it appears to be: A platform that enables AI agents to perform hedge-fund research and strategy development in a controlled, traceable, and auditable way. It provides a structured workflow from market question to backtest to paper trade, with emphasis on reproducibility, observability, and safety boundaries.

What changed: The author describes Ithaca as an operating system for AI hedge-fund research that connects AI agents with financial data, strategy specification, backtesting, and paper trading — all within a single traceable system of record. It is built to address fragmentation in finance tooling and to support human review at key decision points.

Single most important open question: Is there sufficient commercial demand for this specific type of control layer for AI agents in finance, or is the author’s traction limited to early-stage interest from friends and family?

The description states that Ithaca is an operating system for AI hedge-fund research. It enables users to:

• Turn market questions into constrained strategies.
• Benchmark backtests.
• Generate risk traces.
• Run paper trades.

It includes a MCP (Model Context Protocol) gateway, which authenticates tools and scopes access; a JSON Schema-based declarative strategy specification; and a control plane that manages immutable strategy versions, compliance checks, approvals, and the money path. It also supports deterministic backtests, traceable workflows, and a React/Vite UI for observability.

Ithaca is described as a Python platform with a React/Vite UI, using technologies such as FastAPI, Docker, Kubernetes, PostgreSQL, OAuth 2.0, and Interactive Brokers integration.

Inference: The product appears to be a workflow engine for AI agents in finance, designed to make research-to-deployment processes auditable and reproducible.

The author states that Ithaca is positioned as an operating system for AI hedge-fund research. It aims to solve fragmentation in finance tooling by connecting data, strategy authoring, backtesting, and paper trading into a single system of record.

It claims to provide:

• A traceable workflow from prompt to evidence to paper outcome.
• Reproducibility through deterministic backtests and event streams.
• Auditable execution via trace events, artifacts, and human review points.
• Safety boundaries, such as preventing agent code execution, controlling cross-account access, and gating deployment.

The positioning evolved from a general-purpose AI research tool to a control layer for finance-capable agents, with an emphasis on human-in-the-loop decision-making and compliance.

Inference: The author is positioning Ithaca as a specialized control system for AI in finance — not just another chatbot or data tool, but a platform that governs the lifecycle of AI-driven strategies.

The description states that Ithaca targets AI hedge-fund researchers and finance teams using AI agents. It is designed to help users who want to:

• Inspect, replay, and control AI-generated research.
• Ensure compliance and auditability in financial workflows.
• Avoid fragmentation between tools for data, backtesting, and deployment.

It is not described as targeting retail investors or general-purpose AI developers.

Inference: The ICP appears to be finance professionals using AI agents, particularly those working in hedge funds or quantitative research environments where traceability and control are critical.

The description does not include any information about pricing, business model, or monetization strategy. It mentions that the author reached roughly $1,000 MRR through friends, family, and cold outreach — but this is self-reported and unverified.

Inference: The business model is unclear from the provided description. There is no evidence of a formal pricing structure, customer segmentation, or revenue streams beyond the self-reported $1,000 MRR.

Ithaca is built with:

• Python, React/Vite, FastAPI, Docker, Kubernetes, PostgreSQL
• OAuth 2.0, MCP gateway, JSON Schema, Pydantic, Zod
• Interactive Brokers integration
• Server-Sent Events (SSE) for UI updates
• Deterministic backtests with SHA-256 result hashes

It uses a stateless worker model, immutable strategy versions, and an append-only trace event stream.

Inference: The technical stack suggests a modern, scalable platform built for observability and control. It is designed to support reproducibility, auditability, and integration with financial data sources.

The author states that Ithaca reached roughly $1,000 MRR through friends, family, and cold outreach. The strongest signal was not demand for another market chatbot, but rather a desire for an agent workflow they could inspect, replay, and control.

There is no evidence of:

• Customer list
• Revenue history
• Product usage metrics
• Product-market fit validation beyond early-stage interest

Inference: Traction is minimal and self-reported. The author is in the early stages of customer discovery and product hardening.

The description does not mention any direct competitors. It is unclear whether Ithaca competes with:

• General-purpose AI research platforms
• Backtesting tools (e.g., QuantConnect, Backtrader)
• Financial data or trading platforms
• Agent orchestration systems

It positions itself as a control layer for finance-capable agents — not a replacement for existing tools but an enabler of them.

Inference: The competitive landscape is unclear. Ithaca appears to be in a niche space, possibly overlapping with AI agent control systems or financial workflow platforms, but no specific competitors are named.

• Unverified traction: $1,000 MRR is self-reported and unverified.
• No pricing or monetization model is evident.
• Single-founder team: Only one member (Joshua Jerin) is listed.
• Early-stage product: The launch is paper-only, with no live deployments.
• Limited evidence of customer validation beyond early outreach.
• Unclear scalability and adoption potential in a niche market.

Inference: The project is at an early stage, with limited commercial traction or validated demand. It may be vulnerable to misalignment between perceived need and actual market demand.

1. What specific financial workflows are you targeting, and how do they differ from existing tools?
2. Can you provide more details on the $1,000 MRR? Is it recurring, and what is the customer base?
3. How do you plan to scale beyond friends and family for customer acquisition?
4. What are your plans for monetization, and how do you intend to price the platform?
5. How does Ithaca integrate with existing hedge-fund or quantitative research infrastructures?
6. What are the key technical challenges in scaling the traceability and deterministic backtesting features?
7. Are there any partnerships or integrations already in place or planned?