Start with you.
Build for others.

Not yet. Jess said it was focused on cancer survivorship. Parental leave and other situations were discussed as possible extensions, not features already available.

Jess made several iterations over months, keeping the experience relatable and avoiding bulk or complexity that could intimidate the intended audience.

Sam clarified that the One Prompt challenge was ending while Build for Good was about to open. A misleading closing label was discussed; Build for Good was to run for seven days after the session.

Sam's starting heuristic was that something genuinely useful to you may help others too. Jess's example added concrete feedback from a trusted test group and relevant communities. Personal usefulness alone was not presented as a measured market-size study.

Yes. Sam said Office Hours recordings are retained and that the live YouTube VOD can be rewound while the session is still running.

She used communities she already belonged to: LinkedIn breast cancer survivor groups, Facebook groups, and a subreddit. She received useful feedback through those channels.

The AI coach remembers searches and documents created in the app, making subsequent help more personalised without requiring users to repeat the same context.

Verifying the resources and documents, including checking links and whether they applied to her audience. Jess described the process as tedious but necessary.

Jess was considering it. A community member later offered to buy a year's domain registration; the session did not show a completed purchase or connection.

No. Jess described organic distribution through existing circles and interest from different countries. The auto-captions garble the traffic figure, so no numeric growth claim is retained.

Yes. Friends and family checked that it was easy to use, understandable, and not too bulky before broader sharing.

She was still learning. The original article, the natural order of the guide, what she herself would need, and Base44's suggestions shaped priorities. Sam added testing and feedback-board suggestions rather than claiming she already had a formal prioritisation process.

Yes. She named Prompt Cowboy as a free tool for preparing and personalising her initial prompts. The spoken URL is unclear in the captions and is not reproduced as a verified link.

No. She said Base44 felt intuitive enough to build something useful for herself and others without one.

Interactive parts such as document creation and the AI coach needed iteration. Unexpected issues appeared as she built; she worked through them rather than starting with a complete technical roadmap.

Add a feedback feature for bugs, feature ideas, and questions, and save which page a report came from. A Product Hunt- or Reddit-style voting interface can surface priorities. Sam endorsed the idea; no universal automatic bug-fixing guarantee was made.

Open Channels and Roles near the top of Discord and select Spanish, Portuguese, or Japanese. The role unlocks the corresponding moderated community channel.

Sam said there was no tracker, but members could request updates through support tickets.

She was considering other nonprofit projects as well as expanding the existing app. No next product or release date was confirmed.

Yes. She welcomed conversations with organisations that could benefit, including local cancer charities.

Sam said internationalisation could be added with an npm-based prompt and a language selector. He also suggested Spanish and French. This was a proposed implementation for Jess, not a completed translation in the session.

Include English among the supported languages and state that it remains the default. Sam gave this extra instruction after a user reported a translation attempt replacing English text.

Yes, including oncology practices and cancer associations that see a benefit. She also expressed interest in adapting the concept to other return-to-work situations.

Yes. Sam suggested helping a nonprofit or a service such as a food shelter that needs a website, landing page, or CRM but cannot readily fund one. The social benefit, not revenue for the builder, was the point.

Sam discussed paid or donation-supported experiences whose proceeds benefit a cause, including an optional donation banner or a charity-linked game. These were ideas; entries still needed to comply with the published challenge rules.

Look for unmet needs among friends, local organisations, disaster-relief efforts, or people with limited mobility. Sam also supplied a Build for Good idea generator with ten topic buckets.

Sam said prompt credits should not be consumed without builder use and recommended cheaper models and caution with Superagent automations when budgeting. He specifically warned that Fable is costly. The session did not investigate the account's ledger, so no individual billing fault was diagnosed.

Sam said ordinary unused monthly credits do not roll over. He separately said an upgrade carries remaining credits through; that was the session's explanation, not a verified current billing rule.

Sam usually asks for HTML first and then a PDF conversion, but did not resolve this user's RTL case live. He offered to inspect the working HTML and faulty PDF side by side and try a small reproduction.

Some accounts could, as part of a test. Sam directed the user to account settings and did not claim universal access.

Sam said direct transfers would be logistically complicated. He framed building a useful app with your time and credits as the current practical way to contribute, rather than announcing a transfer feature.

Open a support ticket with the billing details. Sam said an actual duplicate charge should be corrected, but the session did not inspect the account or verify that a refund had occurred.

Sam suggested documenting or live-streaming the build because there is an audience for watching people code. It was a content-distribution idea, not a promise of audience growth.