There By — Know When to Leave

Note: This analysis is based entirely on the self-reported, unverified project description provided by the caller. No external corroboration exists for any claims made.

The description states that There By is a tool for NYC subway commuters that answers one question: when do I have to leave so I am not late? It presents:

• One hero leave-by time
• A usual arrival estimate
• A worst-case estimate
• Late risk
• Ranked alternatives when another route is safer

It uses GTFS-Realtime data and a calibrated forecasting model, but the AI models did not invent transit facts or forecast outputs. The product is described as consumer-friendly in language ("Usually," "Worst case," "Late risk") rather than statistical jargon.

The client is an Expo/React Native app with a responsive web experience. The backend is a typed FastAPI service that consumes and archives GTFS-Realtime data, evaluates supported line and transfer paths, and applies uncertainty buffers.

Inference: The product appears to be narrowly focused on a single decision point — when to leave — rather than a general trip planner or map tool.

The author states that the inspiration was rooted in the idea that NYC subway riders do not need another map, but one trustworthy answer before a real deadline. The positioning is narrow and outcome-focused: “when do I have to leave so I am not late?”

There By started as a predictor and calibration foundation, then evolved into an installable beta during Build Week. The evolution included:

• A responsive landing page
• iOS TestFlight distribution
• Expanded station coverage
• Notification hardening
• Observability features

Claim: There By is positioned as a tool that simplifies the commute decision by reducing uncertainty.

The description states that there is one target customer: fixed-schedule NYC commuters. The product is built for people who need to arrive at a specific time and want to know when they must leave, based on live subway conditions.

Inference: The ICP is narrowly defined as NYC subway users with fixed schedules, not general transit users or casual riders.

The description does not state anything about pricing, monetization, or a business model. It only mentions that the beta is now live and that the author is recruiting a small cohort of users to validate leave-by accuracy and notification timing.

Not evidenced: No evidence of revenue, pricing, or monetization strategy.

The product was built using:

• Client: Expo/React Native app with responsive web experience
• Backend: FastAPI in Python
• Data sources: GTFS-Realtime data
• AI tools: Codex (with GPT-5.5 and GPT-5.6)
• Hosting: Railway, Cloudflare R2
• Monitoring: Sentry
• Distribution: TestFlight

The description notes that the engineering challenges included making the last mile trustworthy — broad station coverage, idempotent notifications, observable delivery outcomes, authenticated production services, and recoverable data.

Inference: The technical stack is modern and focused on reliability and observability. The use of AI tools like Codex and GPT-5.6 suggests a rapid development process with AI-assisted engineering.

The description states that the beta is now live, and the author is recruiting a small cohort of fixed-schedule NYC commuters to measure whether they return on real mornings. It also mentions validation of leave-by accuracy and notification timing before broader launch.

Not evidenced: No evidence of actual user adoption, retention, or usage beyond the author’s own testing and beta recruitment efforts.

The description does not mention any competitors. The product is described as answering a specific, narrow question — when to leave — rather than offering a full transit planning experience.

Not evidenced: No competitive landscape or differentiation analysis provided.

• No traction or user data: The project has no evidence of real users or adoption beyond the author’s own beta testing.
• Unproven commercial viability: There is no indication of a business model, pricing, or monetization strategy.
• Limited scope: The product is narrowly focused on NYC subway commuters and does not appear to scale beyond that.
• Self-reported only: All claims are unverified, with no third-party validation or historical data.

Inference: The project appears to be in an early beta stage, lacking commercial traction or evidence of a sustainable business model.

1. What is the actual user retention rate among the fixed-schedule commuters you recruited?
2. How do you plan to monetize this product beyond the current beta phase?
3. Are there any plans to expand beyond NYC subway commuters or include other transit systems?
4. How is the accuracy of the leave-by time validated in real-world conditions?
5. What are the key assumptions behind your forecasting model, and how are they calibrated?