KiP Ai CmS turns your daily cultivation data into one shared, audited record. Every lot you log, every event you track, every transfer you sign off lands in the same database — and when you want intelligence, AI analyses that module's own records on demand.
The more you put in, the more you get back — every engine wired into one shared cultivation system.
Snap a whiteboard, label, sensor display, or dashboard screenshot and server-side Vertex AI (gemini-2.5-flash, no browser key) fills the active module's form — per-card capture pulls one row, bulk whiteboard import returns an editable review grid before commit. Domain-aware (FB/SM/VEG/FIN, red-ink flags, comparators, ratios, blanks as null); quota is charged only on success. Pro and up.
Every irrigation log runs the canonical crop-steering framework: the primary driver is dryback depth (industry-standard (peak − current) / peak × 100, banded shallow → water-stress) and the secondary driver is runoff-EC delta, with five cross-correlation rules combining both — each recommendation cites the rule that fired and projects the expected dryback / runoff % / steer band after the change. K3B0 is the agronomic deep dive (8192-token budget, truncation-retry, structured output); R3DE is the green/yellow/red shift brief. R3DE runs on every tier; K3B0 from Pro.
Record-changing actions across the modules write a timestamped event to a single per-facility audit_log — who, what (module, action, record, before/after) and where from — surfaced through a read-only spreadsheet viewer (the "View Audit Log" button) filterable by module, action, entity, or user email. Capture is durable: the write completes before the response is returned, and a failed write is retried, kept in an outbox and reconciled daily. No client delete path; kept for the life of the account.
On the roadmap, not live yet: logged activity would fill a Seed → Sprout → Canopy → Harvest gauge, unlocking anonymized network benchmarks, strain comparisons, and a network success feed. Activity measures engagement — not accuracy or compliance.
The AI says what it did: it analyses the photo, drafts a summary from your records, or answers questions over your own records. HarVesT (harvest readiness + yield) scores readiness from a trichome photo analysis and states its basis. AI calls are the premium fuel — allowances scale by tier (see pricing), and B0ost / PuLsE / SuRGE top-up bundles are coming soon, pricing announced at launch.
Owners and super-users edit required fields, rename labels (Room → Bay, Runoff EC → Leachate EC), and hide unused modules, with changes propagating facility-wide on next render (facility configuration: Pro and up). Across 17+ irrigation/environment inputs, live helper text surfaces aliases, healthy ranges, and cross-field interpretation (drain vs feed EC, dryback target vs actual).
Any user can pull lot- or strain-based reports from any module in CSV or JSON (PDF from Plus), filtered by lot, strain, room, or date. On Pro+ and above, ExecutivE i i (owner BI dashboard) adds the cross-module facility-wide lot report — every module's records for one lot in a single document.
Invite collaborators to specific lots, facilities, or modules with permission-based access, every record-changing action audited (user, timestamp, facility code, module). Anyone can be invited as a collaborator on other facilities; shared network benchmarks are on the roadmap, not live yet.
The premium compliance differentiators — LP Pro and Enterprise — that make an auditor's day short and your partners part of your chain of custody.
CTLS draft, duty worksheets, custody ledgers, and every sealed document with its hash manifest — one branded, verifiable package, one click. Hand your auditor everything in sixty seconds. LP Pro and Enterprise.
Receivers scan the packing document's QR to verify origin, destination, and seal — and acknowledge receipt — with no login. Your supply chain joins your chain of custody. LP Pro and Enterprise.
An owner's brief, drafted by AI from your records every Monday: what moved, what's due, what's at risk, and one recommendation — before your first coffee. White-label on LP Pro. LP Pro and Enterprise.
Now with Compliance H+ — Health Intelligence™ on every tier: every tier sees its score, and the report's depth grows with tier (full report on LP Pro and Enterprise).
Some features are in beta and require tuning. Be patient and communicate with us — we appreciate you.
Planned — not live yet. On the roadmap: your logging activity fills a gauge, and network visibility is earned. Activity measures engagement, not accuracy or compliance.
Create lots, record production events, transfer inventory, approve releases.
Planned: logged activity would earn Activity Score points.
Planned: Seed → Sprout → Canopy → Harvest, each tier revealing more network data.
Planned: collaborator success feed, strain benchmarks, anonymized regional comparisons.
16 modules: 15 live today. Farm AI (Beta — a preview, not yet backed by live data) completes the suite at launch: 16 of 16 live, targeted Q4 2027. Every one shares one audit trail, one identity, one database. See all modules →
Logging is always free. When you want intelligence, press Analyze — the same class of engine that steers VisioN (daily grow log + photo analysis) runs on that module's own data.
Every entitled module keeps its Log/record action free and unmetered. Asking KiP Quip (plain-language assistant over your own records) about your own logged data is unmetered too, from Pro up — questions are answered from your records, not from a model's memory, and most are computed directly rather than generated. A single Analyze action runs the intelligence engine on that module's own domain data — a single entry (this crop, this work order, this lot) or the full history (yield trends, reorder risk, maintenance backlog, schedule density). That is where an AI call is spent.
Analyze draws from the same shared per-tier AI quota VisioN uses — charged only on success, so a failed or empty run costs nothing. Bring-your-own AI key bypasses quotas entirely; top-up bundles (coming soon, pricing announced at launch) will add calls when you need them. On the free Trial, VisioN earns up to 4 AI calls a month from logging streaks.
Per-module Analyze reasons over that one module, one facility. Cross-module and cross-facility reporting — the true business intelligence — stays exclusive to ExecutivE i i (Pro+ and above). Every module's Analyze can point you there; none of them computes it.
Your modules talk to each other, so your teams don't collide. Conflicts surface where you're working — and resolve themselves when the schedule changes.
CalENdaR (facility calendar + lot schedule) × IpM (pest management + spray records) × EpU (equipment + rooms) × PaLLeT (packing, shipments + chain of custody) cross-checked on every write; conflicts surface on the event, in IpM Alerts, and in the ExecutivE i i attention queue, and resolve themselves when the schedule changes.
In-app bell, per-module email groups, and Slack/Teams webhooks for act-now events, with text alerts coming. Admin sets a facility default and per-person On/Off/Inherit.
In compliance mode every application must state its re-entry interval — never defaulted — and the room is locked for it — PurE Ai refuses work there and CalENdaR flags anything scheduled inside the interval; an MOA-rotation guard works against resistance drift.
16 modules: 15 live today. Farm AI (Beta — a preview, not yet backed by live data) completes the suite at launch: 16 of 16 live, targeted Q4 2027. A few of the modules, in brief — every one is on the modules page in full.
Scouting to enforced action: per-room findings judged against your thresholds, treatments whose re-entry interval must be stated in compliance mode, an MOA-rotation guard, beneficials tied to RacK (inventory + nutrient feeding), and alerts that reach the right team. Standard module slot — Contact.
Packing list to signed release: build manifests or AI-parse a photo/PDF, draw-to-sign release, scheduled shipments straight onto the shared Calendar, inbound deliveries tied to RacK. Available from Pro+, standard module slot — Contact.
Your CTLS monthly report, prepared in minutes: KiP auto-fills every line it can defensibly derive from your records — with provenance — and computes your CRA B300 duty worksheet on the higher-of rule. CSV, JSON, and print exports; audit-logged drafts kept while your account exists — meeting a retention period stays the licence holder's obligation. Every tier.
The cultivation news feed + weekly digest: public industry, regulatory, research and grower feeds fetched every 6 hours and ranked for the strains you grow, each story dated by its source and opening at the source, plus the KiP playbook and a Monday digest. Pro+ and above.
Strain lineage + phenotypes: a strain library, phenotype hunts, and two trees — where a cultivar came from, and what it has actually done in your facility across GarDen, VisioN and HarVesT. Standard module slot — Contact.
Sanitation SOPs + compliance tasks: per-room checklists, completions that record who did the work and when, one open due date per task, and re-entry warnings that reach the person about to walk into a treated room. Standard module slot on any paid tier — Contact.
Sensor-driven automation, in design: industrial PLC and sensor integration (in design), with the control logic authored in KiP. Today it is an in-app preview on simulated data. Enterprise, per room. Pricing announced at launch; early-access terms on request. Request early access →
Farm AI roadmap Beta — a preview, not yet backed by live data
What shipped across the live modules since mid-September.
A Gantt chart per batch, by room and by lot. Propose a change and approve it, with the knock-on rescheduling previewed first; drag a bar to propose a move; move several at once and undo an approval within 10 minutes. Flower rooms are queues, mother / clone / veg rooms are shared, and a stage is scheduled only in a room designated for it. Your own phase lengths and default rooms live in Settings.
A read-only iCal subscription for Outlook, Google or Apple Calendar, the schedule board as CSV, and a one-page print.
The database is exported nightly to a separate, private backup location (Google Cloud, US multi-region) and kept 30 days. Point-in-time recovery covers the last 7 days, to the minute. A full test restore passed on 10 Oct 2026, and completed in under 3 minutes. Owners and admins see every run in Settings → Backups.
Release needs a passing lab certificate, opened and identified at upload and re-checked at release; lab results are append-only. Testing and QA Review cannot be skipped, a lot cannot exceed its harvest, and the server assigns lot numbers.
No shipment is created already in transit; nothing is released beyond the lot's balance; every product line names a lot; Delivered requires In Transit; destination and carrier are sealed, with amendments recorded; freight counts in a fixed list of units; a receipt counts once.
Re-entry warnings reach the person about to enter; completions record who and when; one open due date per task; a past date closes as done, missed, superseded or duplicate.
Upload your plan, set its scale, outline the rooms and place equipment in feet; export a PDF sheet. A photo goes to your records for review, and a replaced plan is kept as a version. Pro and up.
Every module dashboard reads your facility's own live records — the numbers on the screen are the numbers in the database.
Account → Security activity lists your last 100 security events — sign-ins, failed sign-ins, password changes, AI keys, and any change KiP staff made to your account. And an invitation grants nothing until the invited person accepts it.
KiP Quip (plain-language assistant over your own records) summarises any entry across RacK, KpU, EpU, PaLLeT and CalENdaR, and one strain across every module that touched it — never a module your plan excludes.
Mother lots with establishment and regeneration due dates, and the day a batch flips from veg to flower, recorded once.
Audit-by-default records roll straight into Health Canada CTLS preparation and a CRA B300 duty worksheet. Log all month; report in one sitting.
Every record-changing action audited. Every user traceable. Every signature timestamped. Every module writes to the same per-facility audit log.
VisioN captures, GarDen (plants, batches, propagation) client orders, RacK inventory adjustments, KpU (work orders + SOP documents) work orders, Harvest readiness, Calendar cancellations, SafE (lot custody: dry → cure → trim → test → release) lot releases, EpU compliance flips — record-changing actions write a single event to facilities/{fid}/audit_log with module, action, record, user, before/after, where from (user agent, IP, method, path), and server timestamp. The write completes before the response is returned; a failed write is retried, kept in an outbox and reconciled daily. No client delete path; kept for the life of the account.
The "View Audit Log" button on every module opens the shared KipSpreadsheet viewer over the same data — filterable by module, action, entity, or user email. Pull a CSV of every change a specific employee made last quarter, or every SafE lot stage advance in the last 30 days, without leaving the UI.
A SafE release is signed by a PIC or HOS with their name and a confirmation, alongside the destination and notification email — and only once a passing lab certificate is on file. PaLLeT shipments are signed again at release. Nothing leaves the facility without a signature that points back to a named human, and every signature lands in the audit log alongside the stage change.
Order emails follow the exact format Order Complete_LOT - Destination - Date - Initiator so your inbox becomes a searchable second-source audit log. Calendar cancellations send a templated team notification with the original time, room, and reason so nobody shows up to a meeting that was scrubbed.
VisioN is free on the Trial, with one room. Explore every module in the live demo — no signup, no card.