Welcome to Agentic Protocol Tracking thatdrafts the protocol shell from the proposal itself.
The IRB clock is the silent serial step: teams celebrate the award, then discover the review takes ninety days and nobody filed. Risely's agents draft, assemble, and watch every clock, while the board reviews and the PI signs.
Protocols · the lab's clocks
watching
New study
shell drafted from the proposal · filed the day of submission
Renewal · 47 days out
continuing-review packet assembled · the PI signs, and it files
Amendment
new co-investigator · modification drafted the same week
Training
one certification expires next month · reminder already out
The shell drafts from the proposal itself.
The protocol files at submission time, so the ninety-day review runs while the sponsor is still deciding. The award arrives spendable, the postdoc offer goes out, and the first-year timeline stays real.
Two reviews, one calendar
Sponsor review of the proposal
Protocol review · filed the day of submission
The ninety-day review runs while the sponsor is still deciding, so the award that arrives is spendable on day one.
The silent serial step was never the review itself. The step was the months nobody started it.
The renewal assembles before the lapse can start.
At 47 days out, the continuing-review packet is already built from the study's own records. The PI signs, and it files. The lapse mid-collection, which is a compliance incident with a name on it, never happens.
The renewal, 47 days out
assembling
The PI signs, and it files. The study that lapses mid-collection is a compliance incident, and this one never gets the chance.
Amendments happen in the right order.
A new co-investigator gets certified before the modification files, and touches data only after approval. The order sounds obvious, and without the clocks it almost never happens that way.
A person joins the study
Training certified
the new co-investigator completes it first
The modification files
the PI signs the draft the same week
Access granted
they touch data only after approval
The order sounds obvious and almost never happens by itself. The clocks make it the path of least resistance.
The board meets a finished packet.
Risely's agents assemble what the board reads. The review, the judgment, and the approval stay entirely the board's, because a human-subjects decision is nobody's to automate.
Who holds what
Human-subjects review is a compliance surface with legal weight. The agents make the board faster and never make it smaller.
One person joins. The clocks re-run.
A new co-investigator, a protocol change, a training expiry. Any of them moves a clock somewhere, and every one runs the same way.
A co-investigator joins
mid-study, mid-semester
The training check runs first
their certification is verified current
The amendment drafts
the modification, ready the same week
The PI signs, and it files
in the right order for once
The clocks re-set
renewal and training, both visible
What changes, and how it runs.
Awards arrive spendable
The review ran while the sponsor was still deciding.
Studies stop lapsing mid-collection
The renewal packet assembles weeks before the date.
The board's time goes to judgment
Complete packets in, decisions out.
The technical read
Connects read-only to your protocol system, grants system, and training records to start
The board reviews and the PI signs, always
Every action lands on an append-only log
Runs alongside what you have. Nothing rips out.