Mentioned: Building a Brand Monitoring SaaS From Scratch

For as long as I've been a developer, I've wanted to build and ship my own product. Not a client project, not someone else's vision, but something I designed, architected, and owned from the ground up. After 20 years of building software for other companies, I finally decided it was time to build something for myself.

Mentioned is a brand monitoring tool that helps businesses track where and how they appear online. It monitors web mentions, tracks search engine rankings, analyzes backlink profiles, and, increasingly important in 2026 and beyond, detects when a brand appears in AI-generated search results. If someone mentions your company in a blog post, if your search rankings shift, if a new site links to you, or if an AI model starts referencing your brand in its responses, Mentioned surfaces that information in one place.

It's the kind of tool I would have wanted at every company I've worked for, and it's the kind of niche, data-driven SaaS that I believe has real staying power.

Why this product

The brand monitoring space isn't new, but the landscape is shifting in ways that create genuine opportunity. AI-powered search such as tools like Perplexity, Google's AI Overviews, and conversational search more broadly, are changing how businesses get discovered. Traditional SEO monitoring doesn't capture whether your brand is being mentioned in AI-generated answers, and most existing tools haven't caught up yet.

I saw a gap and decided to fill it. Mentioned combines traditional brand monitoring (web mentions, backlinks, search rankings) with AI mention tracking, giving businesses a complete picture of their online presence as the web evolves.

The technical foundation

Mentioned is built with Laravel. It's the framework I know best and the one I believe is the right choice for a solo founder shipping a SaaS product. The architecture centers around several key systems:

Data ingestion via DataForSEO. The platform integrates with multiple DataForSEO API endpoints including Content Analysis, Backlinks, SERP, Domain Metrics, and their AI Optimization and LLM Mentions APIs. Each of these provides a different lens on a brand's online presence, and orchestrating them into a coherent data pipeline was one of the more interesting architectural challenges of the project.

Background job orchestration. Brand monitoring is inherently asynchronous, you're queuing searches, waiting for results, processing data, and storing snapshots over time. I designed a system of background jobs that handle enrichment tasks across these various data sources, with per-tier rate limits to ensure fair usage across the Free, Starter, Pro, and Business subscription plans. Early on I found myself with overlapping jobs that were doing redundant work, so I consolidated them into a cleaner architecture that reduced both complexity and API costs.

Schema design for time-series data. Tracking how a brand's presence changes over time requires thoughtful schema design. I built out tables for search rankings with point-in-time snapshots, AI mentions with their associated sources, and backlink data are all designed to support both current-state queries and historical trend analysis. Getting this right early matters because migrating a time-series schema later, once it's full of data, is painful.

Stripe billing. A tiered subscription system with Stripe, including the billing portal for self-service plan management, webhook handling for subscription lifecycle events, and feature gating based on plan tier. Having done similar work in the past, I knew exactly where the edge cases live: proration, failed payments, plan changes mid-cycle, and built for them from the start.

How AI shaped the development process

This project wasn't "vibe coded." I wrote the architecture, I made the design decisions, I own every line. But AI, specifically Claude, was deeply integrated into my development workflow, and it meaningfully accelerated the process.

Where it helped most:

Thinking through schema design. Some of the most productive conversations I had during this project were with Claude, talking through database schema tradeoffs. "If I structure AI mentions this way, what happens when I need to query for trends over time?" "Should snapshot data live in the same table or a separate one?" These aren't questions the AI answered for me — they're conversations that helped me think more clearly about the tradeoffs before committing to an approach.

API integration boilerplate. Integrating with multiple different DataForSEO endpoints involves a lot of repetitive HTTP client setup, response parsing, and error handling. AI generated the initial scaffolding for each integration quickly, which I then refined and tailored to Mentioned's specific needs. This saved hours of tedious work without sacrificing control.

Test scaffolding. Writing tests for billing webhook handlers, job processing logic, and API response parsing is important but mechanical. AI-generated test scaffolding gave me a starting point that I could expand and customize, rather than writing every assertion from scratch.

Documentation and naming. When you're a solo developer, there's no one to review your naming conventions or point out that your method names are ambiguous. I used AI as a sounding board for naming services, jobs, and database columns as well as other small decisions that compound over a large codebase.

Where I deliberately did not use it:

Architectural decisions. The choice of how to structure the job pipeline, how to separate concerns between data ingestion and presentation, how to design the tier system — these came from my own experience and judgment. AI can suggest patterns, but it can't weigh the specific tradeoffs of your system with your constraints.

Business logic. How Mentioned determines what constitutes a "mention," how it scores relevance, how it decides what data to surface to the user — these are the decisions that define the product. They require product thinking, not code generation.

What this project represents

Mentioned is more than a side project. It's the first real product under Progravity, the company I founded to build and operate my own software products. After two decades of helping other companies ship software, I'm applying everything I've learned about architecture, about product development, about what makes software maintainable over time, into something that's mine.

It's also a practical demonstration of what a solo developer can accomplish in 2026 by combining deep experience with AI-assisted development. I'm not using AI to replace my skills. I'm using it to multiply them; to move faster on the mechanical parts so I can spend more time on the decisions that actually matter.

Mentioned is in active development, after launching it to the public I have no intention of forgetting about it. I hope to continue growing it and making it into something truly unique.