I built a single power scale for people, companies and institutions (what worked, what broke)

Hi all, I'm a solo builder from India. This is my project, and it's still early, so I'd really value honest feedback.

The idea

Forbes ranks billionaires. Wikipedia describes people. But nothing lets you compare a city mayor, a CEO, a ministry and a tech company on the same scale. ApexHierarchy does that: every profile gets one of 8 tiers (Town → City → State → National → Continental → Global, plus two special tiers), with a Low / Mid / High / Top position inside each tier.

What works today

• Search any person, company, institution or country. If it isn't in the index yet, it's researched from public sources on demand (about 25 seconds).
• Six dimensions per profile, each with its own tier: Net Worth, Reach, Power, Influence, Network and Legacy, plus a rising/falling trajectory.
• A connections map: click from Elon Musk to xAI to its own tier, and keep going. Similar-profile suggestions for exploring.
• The Apocalypse Index: a top-10 of entities with the power to reshape world order, plus sourced data on food, water, AI compute and financial plumbing.
• A network graph of each profile's connections.
• Global rankings, net worth rankings, institutions, city rankings across 40+ cities (Mumbai, Chicago, Paris, Tokyo and more), "richest people in" pages for India, the US and China, "most powerful people in" pages for 13 countries, and an A–Z directory of all 777 profiles.
• Every profile lists its sources, marks which are corroborated, and shows a data-confidence score. Anyone can report a mistake, claim their own profile or submit a missing one.
• Free to browse; paid plans (from ₹499/month, or $12/month outside India) add more profile views, PDF export, a watchlist, history and a recent-developments feed.

How a profile is made

Public sources → an LLM writes the text and pulls figures → a formula (not the LLM) assigns the tier. Keeping the tier assignment in code, not model output, was a deliberate choice: the model researches and writes, but never gets to decide "this person is Global tier" on its own judgment. There's also a hard reach-based cap: a profile can't be ranked above what its actual public reach supports, no matter what it claims about itself.

What broke and how it was fixed

• Wrong photos for namesakes (two different people sharing a name got cross-matched).
• Reversed connection labels ("X owns Y" showing up as "Y owns X").
• A security policy that silently blocked Google Analytics from loading.
• A slug-generation bug that mangled accented characters (Düsseldorf became d-sseldorf instead of dusseldorf).
• A dead end for new city-list profiles: the first visitor to a brand-new profile used to only get a real page if they were a paying member — everyone else landed on a "still generating" message that never actually generated anything. Now it generates for the first real visitor regardless of account type.

Numbers

777 profiles, 3,127 connections, 1,648 sources, 40+ cities.

What's next

More city and country lists (each needs 10 fully sourced profiles before it's published), then smaller cities beyond the major hubs. Yearly editions of city lists, the way magazines publish annual rankings. Deeper sourcing and more human verification for the profiles people actually look at.

Honest limitations

Figures are editorial estimates from public sources, not verified financial statements, and profiles are AI-assisted. A few entries are still wrong, and I'm fixing them as they're found. 777 profiles is a small start.

Try it: https://apexhierarchy.com/?utm_source=devto&utm_medium=social&utm_campaign=launch
The Apocalypse Index (the most fun page to start with): https://apexhierarchy.com/apocalypse?utm_source=devto&utm_medium=social&utm_campaign=launch
Methodology: https://apexhierarchy.com/methodology?utm_source=devto&utm_medium=social&utm_campaign=launch

What's missing? Who would you search for first? Anything that looks wrong? I'll answer every comment.