What's New

The latest updates and improvements to Dharma Score.

June 2026

June 15, 2026

A self-service admin management screen, a completed profile-amendments workflow with supporting documents and public annotations, a filterable and exportable analytics dashboard, editorial control over published insights with a new "test a thesis" tool, and faster, pay-for-use scoring infrastructure.

Amendments, Completed

Request Changes to a Specific Section

When you request a correction to your profile, you now choose exactly which part it concerns — the name and photo, biography, character sheet, contributing sources, AI summary, or score breakdown — so the request is unambiguous for you and the reviewer.

Attach Supporting Documents

Attach documents you've already added to your account's verification evidence — no re-uploading — and decide, case by case, whether each one may be shown publicly if your amendment is approved.

Approved Changes Show on Your Profile

Once an administrator approves an amendment, a short, neutral annotation appears at the top of the relevant section of your public score page, so visitors can see the correction or context you provided.

Clear Review States

Every request moves cleanly through pending, under review, approved, or rejected — with a rationale where it matters — and a finished request can't be silently re-opened.

Insights You Control

Publish Only What You Choose

AI-generated insights are now private by default. The team reviews them and publishes only the findings worth surfacing — and published insights stay published even when a new batch is generated, so the public feed no longer disappears on a refresh.

Test a Thesis Against the Data

A new tool evaluates a hypothesis — for example, a claim about how a group is represented — against the full dataset, reporting plainly whether the data supports, partly supports, or refutes it, backed by specific names and numbers.

Analytics Dashboard

The Whole Platform at a Glance

A new analytics view aggregates growth, score distribution, category coverage, demographics with a world map, and pipeline health into a single screen — with totals and their 30-day change computed in one fast pass.

Filter and Export

Filter the entire dashboard by date range, category, and country, then export the filtered view to a single CSV — one clearly labelled section per panel — that opens cleanly in Excel or Google Sheets.

Self-Service Admin Management

Manage Administrators from the Dashboard

The admin area now lists everyone with administrator access — with join and last-sign-in dates — and lets the team promote an existing user by email or revoke access in a single step, no database commands required.

Safe by Design

Only people who have already signed in can be promoted, and the system never lets the last remaining administrator be removed — so the platform can never be left with no one able to manage it.

Under the Hood

Faster, Pay-for-Use Scoring

The scoring worker now runs on Google Cloud Run: it processes multiple scores in parallel, absorbs bursts of demand, and scales to zero when idle, so we pay only for real processing time — with a hard ceiling on worst-case cost.

Visible Recovery

When you request a score, the app now wakes the worker directly and gets a fast acknowledgement. If a job is slow to start, the progress page shows a clear "retrying" state instead of a silent wait.

June 2026

June 11, 2026

A new two-level category system, customizable scoring, evidence-backed profile claims, a public methodology page, and much sharper matching for people who share a name.

Categories & Subcategories

Two-Level Categories

Profiles are now organized by a parent category — Entertainment, Sports, Academia, Health, and more — and an optional subcategory within it, like Actors, Musicians, Soccer, or Basketball. It's a far better fit for comparing the right people.

Filter by Both

Browse Scores and the analytics views now show two rows of category chips — filter first by parent category, then narrow to a subcategory.

Nothing Falls Through

Every existing profile was automatically moved into the new structure, and an "Other" catch-all makes sure no profile drops out of the filters.

Customizable Scoring

Tunable Categories & Weights

Scoring categories, their labels, and their weights can now be adjusted without a software release. The current methodology weights Peer Recognition, Body of Work, Community Impact, Industry Presence, and Public Perception.

Every Score Remembers Its Method

When a score is calculated, the exact categories and weights used are saved with it. If the methodology changes later, older scores still display correctly against the method that produced them.

Versioned History

Each revision of the scoring method is a new version with an effective date and a summary of what changed — an auditable record of how scoring has evolved.

Verify Your Claim with Evidence

Attach Supporting Evidence

When you claim your profile, you can now add a note and attach up to five evidence files. They're stored privately and only ever seen by a reviewing administrator through secure, expiring links.

Build Evidence Anytime

Add a verification note and files to your account before claiming any profile — it's automatically available to the reviewer when you submit a claim, with nothing to re-upload.

Stronger Identity Signal

Linking more accounts and adding evidence strengthens your identity and helps an administrator verify your claim faster.

Careful Human Review

Administrators review your note and files — with image previews — before any claim is accepted. Rejection leaves all of your evidence untouched.

See How Scores Are Calculated

Public "How It Works" Page

A new page explains the whole process in plain language — from submitting a name through AI-assisted research to the final score — and lists every scoring category, what it measures, and what the confidence tiers mean.

Methodology Badge on Every Score

Each score now shows which version of the methodology produced it. Open the badge to see the exact categories and weights used, and a note if a newer version exists.

Accurate Radar Chart

The radar chart now draws the correct shape for however many categories produced a score, with no empty axes — and falls back to a clear bar chart when there are too few categories.

Sharper Person Matching

No More Mixed-Up People

When two people share a name, the platform now remembers a short description of the person you confirmed and uses it to keep the right person's sources and discard the wrong one's.

Smarter Tiebreaking

Distinguishing details like field and birth year separate same-name people automatically, and a small AI tiebreaker resolves the genuinely ambiguous sources instead of guessing.

Better Candidate Ranking

Candidate matches are now ranked by how well they fit the details you provided — country, birth year, field, and your LinkedIn — so the right person leads the list.

More Identity Sources

Matching now draws on additional public sources like Wikidata, OpenAlex, VIAF, and MusicBrainz, with caching to keep lookups fast and costs down.

May 2026

May 20, 2026

Lower-cost, faster sentiment analysis, natural-language search, a live progress page, smarter AI research and quality checks, legal-name display, and the foundation for a browser extension.

Natural-Language Search

Search by Meaning

Browse Scores now understands what you mean, not just exact text. It searches across a profile's name, industry, biography, and the sources behind its score — a query like "Atlanta-based independent filmmakers" now returns real matches.

Match Confidence at a Glance

Each result shows a "% match" chip (or an "exact match" pill), and stage names and nicknames appear as an "aka" line so people surface even when filed under their legal name.

Always Up to Date

Every time a score is recalculated, its search information is refreshed, so results stay current with the latest data.

Live Progress Page

Watch Sources Resolve in Real Time

The progress page now checks each data source one at a time and reveals it nested under its stage — showing "checking…", then "ok", "failed", or "skipped" as it lands.

A Little More Fun

Inspired by a slot machine, the page plays a short tone as each stage begins and a click as each source resolves, with a final flourish when your score is ready.

Smarter AI Research & Quality Checks

Deep Research When Needed

When the usual sources come back thin, an AI research pass runs as an extra source — and its findings stream onto the progress page as they're discovered.

Automatic Quality Review

After a score is computed, an AI pass flags possible concerns — sources that may describe a different person, scores that look inflated, or obviously missing context — for administrators to review.

Faster Disambiguation

When several people share a name, the platform can now research each candidate and auto-confirm a clear match, falling back to asking only when it's genuinely unsure.

Legal-Name Display

Profiles Under Their Real Names

After a score completes, the platform now promotes a person's full legal name to their primary display, keeping the name you searched as an alias so it still resolves. No more profiles filed under just a stage name.

Lower-Cost, Faster Sentiment

More Efficient Analysis

Sentiment analysis now runs through a flat-fee tool first, automatically falling back to the previous service if anything goes wrong — reducing cost with no change to how scores are produced.

Dharma Scanner Foundation

Browser Extension Lookup

A new lookup capability lets a future browser extension flag Dharma-scored people on any web page — matching strictly on exact names and aliases to keep false positives low.

April 2026

April 24, 2026

Major update introducing analytics pages, profile comparison, authentication, profile claiming, and much more.

Smarter, More Accurate Scores

Name Deduplication

We now detect name variations and misspellings automatically, so duplicate profiles are a thing of the past.

Better Source Matching

Sources are now filtered by industry and identity signals like LinkedIn URLs. If you provide a LinkedIn profile, we use it to make sure we're scoring the right person.

Cross-Person Source Filtering

When people share similar names, sources are now correctly attributed to only the right individual — no more data bleed between profiles.

Historical Timeline

Scores now incorporate career history over time, not just a current snapshot. Awards, milestones, and pivotal moments from years past are reflected in your score.

Redesigned Score Details

Cleaner Layout

The score detail page has been redesigned to fit your screen better. No more duplicate sections or unnecessary whitespace. The timeline is always visible.

Character Sheet

Key facts like birth year, nationality, genre, and active years are now displayed in a structured card — separate from the narrative analysis.

Score Change Explanations

When a score is recalculated, you'll see exactly what changed and why — new sources found, sentiment shifts, updated data.

Confidence Notice

The low-confidence data notice now appears at the top of the page where you'll see it immediately, not buried in the middle.

New Analytics & Comparison Tools

Analysis Dashboard

Browse all scored profiles at a glance with industry filters, average scores, and key metrics including representation gap and pay equity ratios.

Side-by-Side Comparison

Select multiple profiles and compare them in a clean, side-by-side table. Scores, categories, and key metrics are aligned for easy comparison.

AI-Powered Insights

A dedicated insights page analyzes the entire dataset to surface patterns and findings you wouldn't spot from individual profiles alone.

Accounts & Authentication

Create an Account

Smart Forms

If you're signed in, the "Get Your Score" form automatically fills in your email — one less thing to type.

Everything Still Open

No features are locked behind login. Accounts are optional and exist so you can claim and manage your own profile.

Claim Your Profile

Multi-Provider Identity Verification

Link your Google, LinkedIn, and X accounts to strengthen your identity claim. More linked accounts means a stronger verification.

One Profile, One Person

Each user can claim exactly one profile. This is enforced at every level to keep the system trustworthy.

Request Corrections & Rebuttals

If something in your profile is wrong, missing, or misleading, submit a structured request. You can file corrections, additions, rebuttals, or removal requests — all reviewed by a human.

Track Your Requests

See the status of every amendment you've submitted: pending, under review, approved, or rejected — with notes from the reviewer.

Under the Hood

Faster Score Calculations

Data sources are now fetched in parallel instead of one at a time, significantly reducing the time it takes to generate a score.

More updates coming soon. Have feedback? Get your score and let us know what you think.