Botsi Raises $1.5M for AI Subscription Pricing

Botsi is moving subscription pricing from a periodic growth experiment into a live decision system. The New York company announced a $1.5M pre-seed round on September 9, 2026, from Telegraph Ventures, Plain Sight Capital, Atlanta Technology Angels, VentureSouth, and angel investors.

The company trains a custom model for each subscription app, then uses behavior and context to select a price, paywall, or offer for an individual user. Instead of asking which single paywall wins on average, Botsi asks which offer is most likely to fit the customer arriving now.

That distinction matters because subscription businesses live inside an awkward trade-off. A higher price can lift revenue per buyer while losing price-sensitive customers, and a lower price can preserve conversion while leaving money behind from customers who would have paid more. Botsi is financing the infrastructure to make many of those decisions at once, then measure whether they actually improve lifetime value.

What Botsi Announced

Botsi's $1.5M pre-seed financing includes Telegraph Ventures, Plain Sight Capital, Atlanta Technology Angels, and VentureSouth, along with unnamed angel investors. The announcement does not identify a lead investor, valuation, prior financing total, or exact allocation of the new capital, so those details remain undisclosed.

The company was founded in 2024 by Jacob Rushfinn, Jason Schubert, and William Schubert. Jacob Rushfinn serves as co-founder and CEO, while Joshua Jarvis is CTO and founding engineer. Botsi says it relocated its headquarters from Charlotte to New York in 2026 and maintains a distributed presence in Georgia, South Carolina, and Texas.

The founding group carries direct experience in subscription growth and mobile monetization. Jacob Rushfinn previously worked on the WordPress mobile app at Automattic and led product marketing at Elevate Labs. Jason Schubert and William Schubert previously built inBrain.ai, while Joshua Jarvis led engineering there before becoming Director of Product and Engineering at Dynata.

Why One Winning Paywall Has Limits

Traditional A/B testing can tell a growth team which version performed best across a population during a defined period. The limitation is embedded in the word "best": the winning result is an average across users with different intent, purchasing power, acquisition sources, devices, and reasons for considering the product.

Botsi's product documentation describes a different operating model. An app sends user and contextual signals to a custom model, which selects among approved prices, paywalls, plans, and offers in real time. The system continues learning as purchases, renewals, cancellations, and traffic patterns create new evidence.

This changes the growth team's job. Pricing is no longer only a campaign that ends when one variant wins; it becomes a production system that must integrate cleanly, preserve controls, explain its measurement, and respond when customer behavior shifts. The upside comes from matching offers more precisely, but the value depends on proving that the model created incremental revenue instead of claiming purchases that would have happened anyway.

The Evidence Behind the Pitch

Botsi says customers including Fitness AI, Sleepiest, and Reading.com have seen revenue and lifetime-value improvements ranging from 15% to 75% since the product launched in Q4 2025. Those figures are company-reported results and should be read as early commercial evidence, not independently audited market performance.

The company's Sleepiest case study provides more detail. Botsi reports that its model went from learning mode to live operation in about three weeks, tested four dynamic paywall variants, and produced a 20% increase in subscriber average revenue per user plus approximately $185K in estimated incremental revenue against a live holdout. A separate Fitness AI case study says the system measured lift against an evenly split control while retraining on richer signals over time.

The control-group language is important because personalized pricing can create noisy success stories. A customer willing to buy at almost any price can make an offer look brilliant even when the model added nothing. A live holdout gives the company and customer a stronger way to ask whether a different decision caused the lift.

The Commercial System Behind the Model

Botsi says integration can begin with a small set of API calls and connect with existing tools such as RevenueCat, Superwall, Adapty, Stripe, Paddle, and web2wave. The company positions itself as a decision layer rather than another paywall builder, which lowers the cost of adoption if the integration and measurement work as described.

The more interesting challenge arrives after deployment. A model selecting different offers must balance revenue optimization with customer trust, privacy, and internal governance. Botsi's Fitness AI case study says its model operates on anonymous signals without capturing personally identifiable information, but every customer still has to decide which inputs and offer ranges are acceptable for its own users.

This is why the financing is larger than a tooling story. Botsi is asking product and growth leaders to put a learning system inside one of their most sensitive commercial moments: the point where a customer decides whether the value is worth the price. App publishers will want more than a lift chart. They will need stable integrations, clean experiments, dependable controls, and a credible answer when the system makes an unexpected choice.

What the Pre-Seed Round Signals

The investor group spans institutional venture firms and regional angel networks, matching a company that began in Charlotte, moved its headquarters to New York, and is selling into a global subscription-app market. The round gives Botsi room to turn early customer evidence into a repeatable operating product, even though the announcement does not specify how the capital will be allocated.

For subscription businesses, the broader signal is that pricing is becoming software infrastructure. Checkout and paywall tools made it easier to present and manage offers; the next layer is deciding which offer belongs in front of which user, and proving the decision improves the economics without damaging the relationship.

Botsi now has to make that decision layer dependable enough to earn a permanent place in the product stack. Each new customer will add data and another opportunity for the model to improve, while also raising the standard for measurement, privacy, and trust that the company must carry as the system keeps learning.