Vol. 01  ·  No. 20 · Wednesday, June 10, 2026
CivicByte.
An independent NYC civic data project · Open source · MIT

What happens when public data reaches people.

People served

142k
Unique visitors · H1 2026

Lookups completed

418k
Across five tools

Records indexed

35M+
From NYC Open Data

Cost to use

Free
No logins · No tracking
The Activity

Where the data went to work.

Through · 23 May 2026
Month What happened Tool Reach
May 2026 Cited in a Bronx Defenders housing-court filing; HPD violation history used to contest an eviction. Building Report 1 case
Apr 2026 Featured in a community newsroom investigation into a 14-building Bronx portfolio with 218 open violations. Landlord Watch 28k reads
Mar 2026 School-budget data presented at a District 9 Community Education Council hearing on Title I allocation. School Budget 1 hearing
Feb 2026 Eviction-history lookups quoted in a Brooklyn tenants' union newsletter circulated to 4,200 households. Eviction Records 4,200 hh
Jan 2026 311 complaint patterns referenced in a community-board testimony on noise enforcement gaps. 311 Lookup 1 board
Dec 2025 Building Report used by mutual-aid organizers to triage heat complaints during the December cold snap. Building Report 11 buildings
In Their Words

What the people using this have told us.

Unsolicited · Verified

The marshal already had the order. I pulled the building's record on my phone in the lobby and showed the judge — fourteen open violations, six of them Class C. We got an adjournment. — Housing-court tenant · The Bronx · April 2026

I'd been asking the district for our school's PTA breakdown for two years. It was already public; I just couldn't find it. Found it on CivicByte in ninety seconds. — Parent association co-chair · Brooklyn · March 2026

We use Landlord Watch as the first pass on any tenant who walks into our clinic. Saves an hour of FOIL requests per case. — Pro-bono housing attorney · Manhattan · February 2026

How We Measure

No accounts. No tracking. Carefully.

Method note

We don't have user accounts and we don't want them. That means we measure carefully: aggregate, non-identifying request counts; referrer patterns when a tool gets shared; and direct outreach from people who tell us how they used the data.

Numbers on this page are rounded to the nearest hundred. Anything quoted as a case, hearing, or filing has been verified with the person involved before publication.

Method note · 2026 H1

See a use we should document, or have one of your own to share?