Deeporax: an AI SaaS that creates and posts social content on autopilot

What was Deeporax built to do?

Running a TikTok page or a product-ad account means producing posts every day, and most of that work repeats: find images, write a hook, lay out the slides, add a call to action, post, then do it again tomorrow. Deeporax exists to take that loop off the creator's plate. The promise on its homepage is simple: automate product ads and TikTok pages, and keep posting while you sleep.

Because it's our own product, there was no client brief. We set the scope ourselves: one place to create content, schedule it, publish it to the networks people actually sell on, and see how each post performed.

What did we build?

The core of the product is the automation. A user picks a slideshow format, checks the hooks, sets a schedule and connects their accounts. From there, Deeporax generates original content and publishes it every day. Around that loop we built:

• A slideshow editor with hook, content and call-to-action slides and live formatting.
• Scheduling and publishing: queue posts ahead or publish straight away, as drafts or live.
• Image sourcing from Pinterest or the user's own collections, plus an inspiration library refreshed daily.
• Consistent AI characters for selfie-style and UGC videos.
• Trending on X, which tracks trends so a user can remix one into a slideshow early, and an AI ads studio where you cast an AI influencer, add a product and get a scripted, rendered ad. Both are marked as new on the site.
• Per-post and per-account analytics for views, likes, shares and engagement.
• A Deeporax MCP server so Claude and other agents can work with the product, on Growth plans and above.
• Free tools alongside the app, including a TikTok hook generator, an image upscaler and an AI voiceover generator.

What were the hard parts?

Posting on autopilot only works if the posts get seen, and new social accounts are fragile. So the product and its help content say it plainly: Deeporax recommends three to seven days of normal browsing to warm up an account before the first post, then one to three posts a day per account. If a page gets zero views, the advice is to pause, browse normally for a few days and come back at one post a day.

The second problem was cost. AI generation costs money every time it runs, so usage is metered in credits. Plan credits reset on the 1st of each month, purchased top-up credits never expire, and the free plan gives a small monthly allowance with no card, so people can try the whole loop before paying.

Connecting several accounts on the same network is awkward too, because the platforms sign you in through the browser. Rather than hide that, the FAQ explains the workaround: sign out of the platform in your browser before connecting the next account.

Which stack did we use, and why?

Deeporax runs on Next.js, Supabase and Cloudflare. Next.js lets the marketing pages, the free tools and the signed-in app share one codebase, with the public pages rendered on the server so search engines can read them. Supabase gives us a managed Postgres database, which suits a product where every automation, scheduled post and credit balance has to stay consistent. Cloudflare sits in front of the site for caching and protection.

What are the results so far?

Deeporax is live and open for sign-ups, with a free plan that needs no card. Direct publishing covers TikTok, Instagram and Facebook, with YouTube and Snapchat marked as coming soon on the site. The site also states that Deeporax is part of the Google for Startups Cloud Program.

We don't quote user numbers for Deeporax here because the site doesn't publish them, and we'd rather point you at the live product than give figures you can't check. If you're planning a SaaS product with AI generation, scheduling or third-party publishing, this is the kind of build our SaaS development team does.