# Helium Automates Mobile Paywall Experiments With Reinforcement Learning and Generative Tools **Author:** Zero G Talent **Published:** 2026-11-07 **Category:** software **Tags:** yc-hiring, reinforcement learning, developer tools, software engineering, product design, golang [Read on Zero G Talent](https://zerogtalent.com/blog/helium-automates-mobile-paywall-experiments-with-reinforcement-learning-and-generative-tools) --- [Helium](https://zerogtalent.com/ai-companies/helium) is building what it calls self-improving software: machine learning systems that automatically generate, deploy, and evaluate user funnel experiments to raise consumer app revenue. Founded in 2024 and backed by more than $5 million from Y Combinator's Summer 2024 batch, GV, Union Square Ventures, Hard Yaka, and leaders from OpenAI, the seven-person team in San Francisco is automating mobile onboarding and paywall optimization for consumer apps with tens of millions of users. Rather than requiring mobile engineering teams to hand-code interface variants or fiddle with manual drag-and-drop dashboard builders, Helium combines an AI editor with automated experimentation infrastructure. Its first customer, a social app with 20 million monthly active users, used the engine to iterate on in-app paywalls and lift subscription conversions. ## Automating the Monetization Funnel Mobile subscription businesses often run into the engineering bottleneck of building and testing paywalls. Designing new variants, coordinating with developers, shipping client updates, and analyzing multivariate experiments can stretch over weeks for a single iteration. Helium replaces this cycle with automated experimentation. On the engineering side, the company's technical stack solves several distinct problems: - **Reinforcement learning for software personalization:** Dynamically assigning subscription pricing, paywall designs, and app experiences to users based on conversion likelihood. - **Automated experimentation engines:** Selecting and running new paywall tests automatically using a combination of behavioral analytics and code-generation models. - **Natural language and design-driven editing:** An AI editor that functions like conversational generative platforms for paywalls, allowing developers to generate layouts through chat prompts or by uploading Figma files directly. - **High-throughput asset delivery:** A Python service built to handle thousands of queries per second (QPS) delivering assets to millions of mobile devices daily, paired with a webview layer engineered to run natively across four mobile client stacks. In an interview on [App Masters](https://www.youtube.com/watch?v=CtQ6XrUrpoo), CEO [Zach Witzel](https://zerogtalent.com/people/zach-witzel) detailed how the platform closes the loop. Helium monitors live user conversion data, generates proactive test suggestions for copy, call-to-action buttons, and hero images, and emails recommendations directly to product teams. Once approved, the changes deploy automatically without custom mobile releases. As Witzel noted, Helium also leverages external payment shifts and web checkout options to test dynamic pricing levels across different geographies and user cohorts. ## Leadership From Uber, Alexa, and Meta Helium's three co-founders bring backgrounds spanning large-scale consumer growth and machine learning infrastructure: - [Zach Witzel](https://zerogtalent.com/people/zach-witzel), CEO, studied computer science at Minerva University. He spent four years at Uber working on Uber Eats growth, surge pricing, and B2B products, managing experiments that drove more than $100 million in revenue. - [Shishir Jakati](https://zerogtalent.com/people/shishir-jakati), CTO, holds a degree in mathematics and computer science from the University of Massachusetts Amherst and previously led early large language model features for Amazon Alexa. - [Anish Doshi](https://zerogtalent.com/people/anish-doshi), Chief Scientist, completed a master's in computer science at Touro University California and previously built machine learning systems at Meta to automate ad experimentation on Instagram, lifting conversion by 10%. According to company materials, three out of four team members have previously founded companies, and all three co-founders hold degrees in machine learning disciplines. ## Scaling the Founding Engineering and Design Team To handle traffic expanding across top consumer mobile apps, Helium is hiring for three onsite founding roles at its San Francisco office. The team offers covered meals, transport, gym benefits, and early access to unreleased AI models. The engineering priorities center on shifting high-scale infrastructure into compiled services and refining the generative design workflows: - [Founding Sr Full-Stack Engineer](https://zerogtalent.com/ai-jobs/helium/founding-sr-full-stack-engineer-88261) ($150,000–$250,000, Mid): Responsible for large components of the AI Experience Generation product. The engineer will build features in a Next.js front-end application, deploy scalable Golang backend code, optimize and load test high-concurrency servers, and design data systems alongside ML training pipelines for automated experimentation. - [Founding Full-Stack Engineer](https://zerogtalent.com/ai-jobs/helium/founding-full-stack-engineer-88276) ($125,000–$200,000, Entry): Works alongside senior staff to build Next.js interfaces, deploy Golang infrastructure, run server load testing, and develop the underlying ML infrastructure supporting automated tests. - [Founding Design Engineer](https://zerogtalent.com/ai-jobs/helium/founding-design-engineer-88274) ($150,000–$200,000, Entry): Owns design across the entire AI experience product. The work involves designing and implementing end-to-end flows using Helium's paywall editor, collaborating with engineers on UX challenges, establishing the visual identity and design system, and conducting weekly customer interviews. ## Open Roles at Helium All three positions are based full-time onsite at Helium's San Francisco headquarters. | Role | Seniority | Location | Work Mode | Salary | |---|---|---|---|---| | [Founding Sr Full-Stack Engineer](https://zerogtalent.com/ai-jobs/helium/founding-sr-full-stack-engineer-88261) | Mid | San Francisco, CA, US | Onsite | $150,000 – $250,000 | | [Founding Full-Stack Engineer](https://zerogtalent.com/ai-jobs/helium/founding-full-stack-engineer-88276) | Entry | San Francisco, CA, US | Onsite | $125,000 – $200,000 | | [Founding Design Engineer](https://zerogtalent.com/ai-jobs/helium/founding-design-engineer-88274) | Entry | San Francisco, CA, US | Onsite | $150,000 – $200,000 | [See all open roles at Helium](https://zerogtalent.com/ai-companies/helium) --- *Source: [Zero G Talent](https://zerogtalent.com), the job board for space, defense, AI, robotics and frontier-tech companies.*