↓ scroll or → arrow keys
01 / 09
Field-to-office · built for GCs

Know where you stand.
Then know what's next.

One running log for the field, the PM, and management. Capture a conflict or an RFI once and it's on the board, rolled up, and legible to management without anyone re-keying it. All from where it happened.

changelog.systems
Why builders use it

Built in your language, not bent to fit.

Capture in the field

Conduit run clashes the structural beam at grid C-4? Log it once, on the job, in the moment not back at the trailer three hours later.

It rolls up by itself

Conflict → RFI → the status-mix, the completions trend, the per-job load. Management sees which job is busy without chasing anyone for a report.

Built in your language

complete · in progress · delayed · conflict · RFI · submittal. Status tags a builder already thinks in — conflict reads red, RFI reads purple, so the entries that matter get their own slices.

Built for GCs running multiple jobs where the field-to-office handoff is the leak — occupied-hospital fit-outs, lab build-outs, renovations, anywhere "captured once, legible to management, nobody re-keying" is the win.

The problem

Status lives in people's heads.

On a real job, progress isn't linear. Work gets interrupted, re-prioritized, blocked on someone else, sent back for revision. By the time anyone asks "where are we?", the why is gone — and the only record is whatever someone remembers in the moment.

Schedules lie

Gantt charts show the plan. They don't show the six interruptions that moved the date.

Reasons evaporate

"Why is this late?" gets answered from memory, days later, defensively.

Patterns stay hidden

No one can see that the same blocker keeps recurring, because nobody's logging it.

The product

A field notebook, not a dashboard.

Log a status change in under a minute, from the floor. Five fields, append-only, plain-text first. It feels like jotting a note — but every entry is structured data the moment it's written.

  • Entries input in under a minute — zero friction between thought and action
  • Status tags that match real work — complete, in progress, delayed, conflict, RFI, submittal
  • Per-job filtering & live "active now" view — what's open across every project
  • Close a job, auto-logs the closing entry — the record writes itself
  • Write-once, append-only — history is a record, not a draft
  • Local-first, no account required — it works before it's connected to anything
  • Plain-text markdown underneath — your data, portable, nothing proprietary
  • Export to markdown anytime — and every integration is a view on top
Why it's defensible

Five fields and an auto-stamped date. Every feature is a view on top.

Reports, calendars, Gantt exports, pattern analysis, AI summaries — none of them need a translation layer, because there's nothing proprietary to translate. The simplicity of the data model is what makes the integration potential real.

date
auto-stamped
·
job
which project
·
status
state of the work
·
area of work
where on the job
·
change
what happened
·
reason
why

Job, status, area of work, change, reason — captured from the floor in one line, with the date stamping itself. The full status update a PM would otherwise chase across a dozen conversations.

Live data · real production use

Already running on real work.

A demonstration dataset — eight weeks of activity across 32 jobs, each entry input in under a minute. This is the shape of the record CHANGELOG produces.

Entries logged
110
Jobs tracked
32
Completions
46
Active weeks
8
The dashboard

Completed jobs over time.

The same six-column log, rendered as the view a manager actually wants: cumulative completions climbing, with weekly activity underneath.

The dashboard

Where the work actually goes.

Completions are less than half the story. Active work, delays, conflicts, and open RFIs are the hidden cost of every project — and for the first time, they're measured instead of remembered.

  • Delays & conflicts logged with reasons — so the same recurring blocker can finally be named.
  • Active work & open RFIs, always visible — what's open across every project, right now.
  • Completions tracked to close — with the closing entry auto-generated.
Where it goes

Ship the log. Then layer on top.

Layer 1 · shipping

The log

  • Under-a-minute entry
  • Active-jobs view
  • Per-job filtering
  • Markdown export
  • Local-first
In daily field use
Layer 2 · integrations

Pipe it anywhere

  • Calendar sync (Google / Apple / Outlook)
  • Notion · Smartsheet · Monday
  • Gantt-readable export
  • Webhooks & CSV
  • Obsidian vault sync
Same data model. No translation layer.
Layer 3 · intelligence

The payoff

  • Voice entry on the go
  • Stale-task notifications
  • Weekly accomplishment digest
  • Pattern recognition
  • Cross-team project rollups
"You log most delays on Fridays."

"Help people focus on what needs to get done
by understanding how it got done."— CHANGELOG · the mission, v0.1

changelog.systems
Try It →