How it works:
Paste in a band name, Reddit comment, Spotify link, or any text about a band or artist
O.B.S.C.O.R.E. (Obscure Band Signal of Cultural Oversaturation, Reach & Exposure) extracts band names using an LLM to figure out the context
The system pulls data from internet cultural hot-spots to measure cultural footprint
Each band receives a numerical obscurity rating based on streaming stats, search data, and online presence
Settle whether your music taste is truly as underground as claimed
Most importantly: It works as an MCP – meaning it can be dropped into your AI workflow, from simple text editors to (soon) complex enterprise systems (contact us if you’re interested)
What’s an MCP and why should I care?
Think of MCPs as “plug-and-play skills” for AI systems. Instead of building custom integrations for every tool you use, an MCP creates a standardized way for AI to access specialized capabilities.
For businesses, this means:
👉 You can add new AI capabilities without rebuilding your tech stack
👉 Your team can use powerful tools through simple commands in their existing AI workflows
👉 Specialized functions can be shared across departments without technical friction
Our band obscurity tool is just one example. The same approach works for analyzing financial data, processing legal documents, or automating customer service tasks.
Things we learned from the thing:
𝗟𝗟𝗠𝘀 + 𝗱𝗮𝘁𝗮 𝗔𝗣𝗜𝘀 = 𝗮 𝗽𝗲𝗿𝗳𝗲𝗰𝘁 𝗺𝗮𝘁𝗰𝗵. AI excels at extracting meaning from messy input. We wanted to build something that doesn’t require a user to perfectly spell “The Tony Danza Tapdance Extravaganza.” We wanted to create a flexible tool that can read a whole Reddit comment or figure out what you’re trying to say even if you’re using voice commands to input a band like “!!!” So regex is out, LLMs process the data, find a canonical band name. Then external APIs measure cultural footprint, counting search hits and checking presence across platforms. Together, they turn messy human language into structured data without complex parsing rules. It’s a pattern that applies to everything from sales conversations to support tickets.
𝗦𝗰𝗼𝗿𝗶𝗻𝗴 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝘀 𝗯𝗮𝗹𝗮𝗻𝗰𝗶𝗻𝗴 𝗺𝘂𝗹𝘁𝗶𝗽𝗹𝗲 𝘀𝗶𝗴𝗻𝗮𝗹𝘀. Some bands have huge monthly listeners, but zero Wikipedia presence. Others have deep cult followings, but barely any Google footprint. The final score blends search hits, YouTube presence, Wikipedia entry length, and Spotify monthly listeners. Especially when measuring complex info (customer satisfaction, product quality, team performance), the magic isn’t in individual metrics, but in how you weight and combine them.
𝗚𝗹𝘂𝗲 𝗰𝗼𝗱𝗲 𝗶𝘀 𝘁𝗵𝗲 𝗿𝗲𝗮𝗹 𝗽𝗿𝗼𝗱𝘂𝗰𝘁. The secret isn’t fancy algorithms. It’s the unglamorous code that retries failed fetches, normalizes weird API responses, and filters out generic words from band names. Without this glue, it’s just broken pipes. This represents 80% of what makes our tool work, but it’s invisible. That (mostly hidden) integration layer that connects systems often delivers more value than any single component.
𝗠𝗖𝗣𝘀 𝗺𝗮𝗸𝗲 𝗔𝗜 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗳𝗼𝗿 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀. O.B.S.C.O.R.E. (Obscure Band Signal of Cultural Oversaturation, Reach & Exposure) started as a fun project, but building it as an MCP—O.B.S.C.O.R.E. M.C.P. (Obscure Band Signal of Cultural Oversaturation, Reach & Exposure Model Context Protocol)—means anyone can now use a simple command to access our entire scoring system without touching any code. A marketer typing /analyze_band_obscurity bandName="Braid" in their chat tool and gets the same results as a developer making complex API calls. Turning specialized knowledge into accessible commands is how businesses will actually integrate AI into everyday workflows. (I know, it’s not well supported yet—but soon!)
