5 OS SaaS ideas to launch in 2026

SAS Opportunities Beyond AI: A Deep Dive into Bare Metal & Open Source

Key Concepts:

• SAS (Software as a Service): Delivering software over the internet, typically on a subscription basis.
• Bare Metal Infrastructure: Utilizing physical servers directly, bypassing cloud providers.
• Open Source Software: Software with publicly accessible source code, allowing for modification and self-hosting.
• Vendor Lock-in: Dependence on a specific vendor’s platform, making switching difficult and costly.
• Agentic AI: AI systems designed to autonomously perform tasks and achieve goals.
• Open Core Model: A business model where a core product is open-source, with premium features offered under a paid license.
• Usage-Based Pricing: Charging customers based on their consumption of a service.
• Lifetime License: A one-time purchase granting perpetual access to software.

I. The Shift Away from Big Tech & Cloud Dependence

The core premise discussed is a growing trend among SaaS founders to move away from reliance on large technology companies and cloud infrastructure. This is driven by concerns over diminishing margins and a loss of control over data and operations. The speaker, Simon Horberg, highlights a “massive migration from the cloud back to bare metal infrastructure and open-source software,” positioning this movement as a significant opportunity for new SaaS ventures. He frames this as a reaction to the convenience of big tech coming at a cost. This shift is not about competing in the AI space, but capitalizing on the resulting demand for alternative infrastructure and tools.

II. Open-Source Affiliate Marketing Tool

This idea stems from a personal pain point experienced while launching the Founder Stack partner program. Existing affiliate marketing tools require connecting to Stripe, granting them access to sensitive financial data (revenue, customer emails, transaction history). This raised privacy concerns, leading the team to build a custom solution.

• Problem: Lack of privacy and transparency in existing affiliate marketing tools. Concerns about sharing sensitive financial data with third parties. Difficulty verifying payout accuracy and attribution logic.
• Solution: An open-source affiliate marketing tool that can be self-hosted, keeping Stripe keys and customer data secure. Publicly accessible code for transparent attribution logic.
• Monetization:
  • Free Community Edition: Fully open-source, encouraging adoption and establishing a standard.
  • Pro Lifetime License: A paid version with advanced features (multi-team support, advanced analytics).
  • Done-For-You Service: A merchant of record service handling tax compliance and affiliate payouts, running on the user’s server.

III. Agentic QA Tester

Inspired by a tweet discussing the diminishing need to understand AI-generated code, this idea proposes automating quality assurance (QA) testing using AI. The speaker notes that having AI review AI-generated code improves quality.

• Problem: Maintaining comprehensive Cypress or Playwright tests is time-consuming and prone to breakage with code changes. Manual QA is necessary but inefficient.
• Solution: An “agentic” QA tester that uses computer vision and a headless browser to simulate human user interactions, testing specific workflows (e.g., “Go to the pricing page, select the pro plan, and make sure the checkout flow works”).
• Monetization:
  • Usage-Based Pricing: Charging per test run, agent minute, or AI credits.
  • Enterprise Plan: Running the agent inside the user’s Virtual Private Cloud (VPC) for enhanced security and compliance.

IV. Shadow Database Tool

This idea addresses the risk of vendor lock-in associated with popular no-code tools like Notion and Airtable. While these tools offer excellent user experiences, they retain complete ownership of user data.

• Problem: Vendor lock-in – the risk of losing access to critical business data if a SaaS provider increases prices, changes terms, or suspends an account.
• Solution: A data liberation engine that syncs data from no-code platforms in real-time to a standard PostgreSQL database hosted on the user’s server. This creates a live, read-only mirror of the data.
• Monetization:
  • Open Core Model: The core sync engine is free and open-source to build trust.
  • Paid Connectors: Charging for connectors to enterprise platforms like Salesforce, SAP, and Oracle.
  • Managed Cloud Version: Hosting the shadow database for a monthly subscription.

V. Self-Hosted AI Media Studio

This idea draws a parallel between the text-based AI world (with tools like TypingMind) and the AI image/video generation space. TypingMind allows users to connect their own API keys, paying wholesale prices for AI inference.

• Problem: AI image/video platforms (Midjourney, Runway) are closed ecosystems. Aggregators (OpenArt) resell compute with a markup.
• Solution: A self-hostable UI that sits on top of AI providers like Replicate or Fell AI, allowing users to chain together models (Flux, Nano Banana, cling, Pixieverse, Sora) using their own API keys.
• Monetization:
  • Lifetime License: A one-time purchase for the software interface. Users pay AI providers directly for generation.

VI. Self-Hostable YouTube Clone

This idea is motivated by recent experiences of creators (like Pat Walls of Starter Story) facing unexpected account suspensions and content flagging on YouTube. It highlights the risk of building a business on “rented land.”

• Problem: Dependence on centralized platforms like YouTube, which can arbitrarily suspend accounts and delete content.
• Solution: A self-hostable platform that replicates YouTube’s functionality (video hosting, streaming, UI elements) allowing creators to maintain control over their content.
• Monetization:
  • Free Community Edition: For basic video hosting.
  • Paid Membership Features: Allowing creators to create a Patreon-like subscription service with paywalled content.

Logical Connections & Synthesis:

The video presents a cohesive argument: the increasing dissatisfaction with big tech and cloud dependence is creating a significant opportunity for SaaS founders. Each idea directly addresses a pain point arising from this trend, offering solutions that prioritize data privacy, control, and transparency. The emphasis on open-source, self-hosting, and alternative monetization models (lifetime licenses, usage-based pricing) reflects a desire to build sustainable, independent businesses. The final idea, the YouTube clone, underscores the broader theme of platform risk and the need for creators to own their audience relationships.

Notable Quote:

“The market is begging for a solution like this.” – Simon Horberg, referring to the open-source affiliate marketing tool.

Conclusion:

Simon Horberg identifies five promising SaaS opportunities driven by a growing movement towards greater control and independence from large technology companies. These ideas – an open-source affiliate tool, an agentic QA tester, a shadow database, a self-hosted AI media studio, and a YouTube clone – all address specific pain points related to data privacy, vendor lock-in, and platform risk. The emphasis on open-source, self-hosting, and alternative monetization strategies positions these ventures as potentially lucrative and sustainable options for SaaS founders in 2026 and beyond. The call to action is clear: identify a problem, leverage the resources available (like the linked video on building a niche AI SAS), and start building.