Skip to content
Provenza
Agentic lab notebook

Reproducible by default.

Provenza is the agentic lab notebook that captures every protocol, reagent, and deviation as you work, so any experiment can be rerun exactly.

Free for solo researchers. No card.

her2-western.protocol v9

“Incubate primary antibody 1:1000 in 5% BSA overnight at 4C”

05

Block

5% milk / TBST60 min
06

Primary antibody

anti-HER2lot AB-2210-071:100016 h @ 4C
07

Secondary antibody

HRP anti-rabbit1:5000
92

Reproducibility score

Ready to rerun

1 deviation vs v8

Trusted by 60 labs to version 8,000+ experiments. (illustrative)

Halden Bio
Meridian Labs
Northcell
Assay Works
Verinox
Coldharbor
The problem

Results that will not come back

An estimated half of preclinical results fail to reproduce, and the record of how each experiment was actually run is usually the first thing to go missing. Scientists lose about a full day a week to notebook admin, and the details that matter, the lot, the incubation time, the one changed step, live in Slack, on paper, or in someone's memory.

~50%of preclinical results fail to reproduce
1 daya week lost per scientist to notebook admin
3 placesa protocol hides: Slack, paper, memory
How it works

Capture once. Version everything. Rerun exactly.

Provenza sits between the bench and the record, so the protocol is built as the work happens instead of reconstructed from memory afterward.

  1. 01

    Log your protocol once

    Speak or type what you are doing at the bench, in plain language, and Provenza captures it in real time instead of asking you to write it up later.

  2. 02

    It structures and versions it

    Provenza parses your entry into ordered steps, reagents, lots, and conditions, then saves a new version every time any of them changes.

  3. 03

    Rerun exactly when results drift

    When an outcome shifts, Provenza points to the precise deviation and rebuilds a clean, rerunnable protocol you can hand to anyone.

Features

Everything an experiment needs to be rerun

Five parts of one system: capture the work, track the inputs, catch the drift, and hand off a record anyone can reproduce.

Voice and text protocol capture

Log any protocol at the bench by typing or speaking, and Provenza records it as you work instead of after the fact.

Dictate hands-free while gloved. Transcripts are structured into steps in seconds.

Automatic reagent and lot tracking

Provenza pulls every reagent, catalog number, and lot into the run record the moment you mention it.

Ties each result to the exact lot used, so a bad batch is traceable across every experiment it touched.

Deviation detection across runs

When a run differs from the reference protocol, Provenza compares the two and marks the exact step, reagent, or condition that changed.

Flags timing, temperature, and concentration drift down to the individual step.

Reproducibility score per experiment

Every experiment gets a reproducibility score based on how completely its protocol, reagents, and conditions are captured.

One 0 to 100 score shows what is missing before another scientist can rerun the work.

Export to ELN and regulatory formats

Export any versioned protocol to your existing ELN or to regulatory-ready formats without rewriting it.

One-click export to ELN, PDF, and structured formats for audit and filing.

Reproducibility diff

See exactly what changed between two runs

This is a real diff of two Western blot runs. Provenza lines them up, flags every silent deviation, and rebuilds a protocol pinned to the reference. Try rebuilding it.

Western blot: HER2 detection in SK-BR-3 lysate

Western blot (SDS-PAGE + wet transfer, chemiluminescent detection)

Rerun vs reference

WB-2026-0627-B

2026-06-27

01Thaw lysate1 changed

Thaw SK-BR-3 whole-cell lysate on ice; second freeze-thaw cycle.Thaw SK-BR-3 whole-cell lysate on ice; third freeze-thaw cycle.

Reagent:
SK-BR-3 lysate (2.4 mg/mL)
Lot:
LYS-SKBR3-2405
Amount:
25 ug/lane
Temp:
4 C
Duration:
20 min
02Denature and load

Mix with 4x Laemmli + DTT, heat, load alongside ladder.

Reagent:
4x Laemmli buffer + 50 mM DTT
Lot:
LMM-0322
Amount:
20 uL/well
Temp:
95 C
Duration:
5 min
03SDS-PAGE

Resolve on 4-12% Bis-Tris gel in MOPS running buffer.

Reagent:
4-12% Bis-Tris gel
Lot:
GEL-BT412-1911
Amount:
MOPS 1x
Temp:
RT
Duration:
50 min at 180 V
04Wet transfer

Transfer to 0.45 um PVDF, methanol-activated, in Towbin buffer.

Reagent:
PVDF membrane 0.45 um
Lot:
PVDF-045-2208
Amount:
Towbin + 20% MeOH
Temp:
4 C
Duration:
90 min at 100 V
05Block1 changed

Block membrane in 5% non-fat milk in TBST on rocker.

Reagent:
5% non-fat milk / TBST
Lot:
MILK-BLK-2401
Amount:
20 mL
Temp:
RT
Duration:
60 min30 min
06Primary antibody5 changed

Incubate with anti-HER2 rabbit mAb, overnight at 4 C.Incubate with anti-HER2 rabbit mAb, 2 h at room temperature.

Reagent:
Anti-HER2 (29D8) rabbit mAb
Lot:
AB-HER2-2210-07AB-HER2-2604-11
Amount:
1:1000 in 5% BSA/TBST1:750 in 5% BSA/TBST
Temp:
4 CRT
Duration:
16 h2 h
07Secondary antibody

Wash 3x TBST, then HRP anti-rabbit IgG at RT.

Reagent:
HRP goat anti-rabbit IgG
Lot:
AB-HRP-GAR-2312
Amount:
1:5000 in TBST
Temp:
RT
Duration:
60 min
08Develop

Wash 3x TBST, apply ECL, image at fixed exposure.

Reagent:
ECL substrate (standard)
Lot:
ECL-STD-2405
Amount:
1 mL/membrane
Temp:
RT
Duration:
2 min substrate; 30 s exposure

Sample data, illustrative. Interactive: toggle highlighting or rebuild the protocol.

From the bench

Scientists who stopped losing weeks to drift

Illustrative quotes from the kind of teams Provenza is built for.

Provenza caught a reagent lot change that explained three weeks of drift we had blamed on our cells. Now every run carries its own record, and rerunning a colleague's experiment takes an afternoon instead of a fortnight.
ES

Dr. Elena Sato

Computational Biology, Boston

We version 400 protocols across four teams, and I no longer chase people for what they actually did at the bench. Onboarding a new hire went from reading scattered docs to reading one clean, versioned protocol.
RM

Rahul Menon

Lab Operations, San Diego

My students capture protocols by voice mid-experiment, so nothing gets reconstructed from memory the night before a meeting. When a result does not replicate, we can see the exact step that changed instead of arguing about it.
PN

Dr. Priya Nair

Principal Investigator, Molecular Biology, University of Michigan

Names and quotes are illustrative.

Pricing

Start free. Upgrade when the lab grows.

Free forever for solo researchers. One flat seat price for teams. Custom terms when security and compliance matter.

Researcher

For a single scientist trying Provenza on real work.

$0

Free forever. No card.

Start free
  • 1 user
  • Up to 25 experiments
  • Voice and text protocol capture
  • Structured steps, reagents, and conditions
  • Full version history
  • Reproducibility score per experiment
  • Community support
Recommended

Lab

For labs and teams that version everything.

$41/ user / mo

billed annually, per user

Start free trial
  • Everything in Researcher
  • Unlimited experiments
  • Deviation detection across runs
  • Automatic reagent and lot tracking
  • Team workspaces and roles
  • Export to ELN and regulatory formats
  • Priority email support

Org

For organizations with security and compliance needs.

Custom

Custom pricing

Contact sales
  • Everything in Lab
  • SSO and SAML
  • Full audit log of reads and changes
  • On-prem deployment option
  • Module-level licensing
  • Custom DPA and security review
  • Dedicated support

Need SSO, audit, or on-prem? Compare all plans.

FAQ

The questions labs ask first

Straight answers on what Provenza is, how it handles your data, and where it fits.

Reproducible by default, from run one.

Start capturing protocols the way you actually work, and make every experiment rerunnable without adding a minute of paperwork.

Free for solo researchers. No card.