NEW Query Predictiv's models from Claude, or any AI assistant, via our MCP server. See how it works →
Demand Intelligence That Drives Action

Stop choosing between
stockouts and dead inventory.

Most planning tools buffer with a rule of thumb that under-covers your longest lead times. Predictiv models the full demand distribution, then hands your team the order to place.

Forecast Optimize Simulate Act
app.getpredictiv.com/actions
Action Cards · Today
3 open
!
Reorder · SKU-2048 Hydro Bottle
Projected stockout in 12 days. Order 480 units by Aug 20.
You actTurn into PO →
Overstock · SKU-1177 Trail Mug
60+ days of cover on hand. Pause the next PO, free cash.
You actReview →
Auto-run · Reorder point raised on SKU-3310
Demand rising. Agent adjusted the policy inside your rules.
Agent actedView log →
Runs on the stack you already have
Shopify
Storefront
Amazon
Marketplace
NetSuite
ERP
Cin7
Inventory
ShipBob
3PL
QuickBooks
Finance
Pre-built connectors plus flat-file import. No engineering required, no rip-and-replace. See all integrations →
The Problem

Planning hasn't kept up
with the business.

Four failure modes show up in almost every brand we talk to, and they compound each other.

01
Forecasts Go Stale Fast
Spreadsheets and last-year averages miss shifting demand, promotions, and seasonality.
02
Stockouts and Overstock
The same catalog is starving best-sellers while cash sits frozen in slow movers.
03
S&IOP Is Slow and Manual
Consensus planning takes weeks of stitching exports together. Decisions arrive late.
04
Insight Lives Far From Action
Dashboards describe the problem; someone still has to figure out what to actually do.
Failure One
Lost sales
Stock out and demand walks to a competitor, or away entirely. Service levels and revenue both take the hit.
Failure Two
Trapped cash
Over-buy and working capital freezes on the shelf: carrying cost, markdowns, and obsolescence.
Root cause: demand intelligence that never reaches the decision. Predictiv closes that gap.
The Platform

One platform,
from signal to action.

Forecast, optimize, simulate, act, with AI and agents running through every step. Every turn of the loop compounds: better forecasts sharpen policies, optimized policies feed cleaner simulations, simulations produce the right action cards, and outcomes retrain the models.

STEP 01
Forecast
Leading-edge demand models. Probabilistic, rolling, every SKU and channel.
STEP 02
Optimize
Right stock, right place, right time. A buying policy for every SKU.
STEP 03
Simulate
Test inventory and spend what-ifs before a single PO goes out.
STEP 04
Act
Recommended moves as action cards. Run them, or automate them.
AI, end to end
Query your models in plain language, and let planning agents run scenarios and trigger the actions inside your guardrails.
Capabilities

What the platform
actually does.

Capability 01 · Forecasting
Probabilistic forecasting, not a point guess.
Most tools give you a single number and a rule of thumb. Predictiv models the full distribution of demand, so the service level you target is the one you get.
Probabilistic, not a point
Median with 25 to 75 and 5 to 95 bands. Plan to a service level, not a guess.
18-month rolling outlook
Refreshes as new data lands. Never a stale quarterly snapshot.
Every SKU and channel
Drill from total demand down to a single item on a single channel.
THE MOAT · STOCKOUT PROBABILITY
We model the probability of a stockout.
A full demand distribution
Every SKU is a probability curve, P5 to P95, not a single number.
Your real exposure window
We roll demand across lead time plus review: the window where a stockout can happen.
Calibrated, not assumed
Coverage is calibrated from your own data, so a 95% target means about 5% real risk. Not a √-of-time approximation that under-covers long lead times.
Capability 02 · Promotions
Plan promotions, not surprises.
Model a promo or marketing push and see the demand lift, the inventory impact, and the buys it forces, before you commit.
1
Model the lift
Add a promo or campaign; the forecast reflects the demand spike it will create.
2
See the inventory impact
Projections update immediately. Where you'll run short, where you'll overshoot.
3
Get the pre-buys
The ordering actions to cover the event land as cards, timed to each lead time.
HOLIDAY PROMO
Nov 20 – Dec 2 · Top category
ACTIONS REQUIRED
Pre-buy SKU-2048
+400 units, arrive by Nov 15
Bring SKU-3310 early
Split across two vendors
Hold SKU-1177
Already covered, no action
Capability 03 · Lifecycle
From launch to last unit.
New items and dying items are where planning hurts most. Predictiv handles both ends of the lifecycle.
Launch Ramp Mature Decline Successor
NEW PRODUCTS
Forecast with no history
Uses attributes and analog items to seed demand from day one.
Plan the launch buy
Ramp-aware initial orders so you don't over- or under-commit.
Watch the ramp
Projections and cards adjust as the first real sales arrive.
END OF LIFE
Discontinue cleanly
Successor / predecessor chains carry demand to the replacement.
Sell down to zero
Draw stock to nothing without a final over-buy or write-off.
No dead weight
Dead SKUs stop pulling reorders and skewing the plan.
Capability 04 · Inventory Optimization
Right stock, right place, right time.
Turn the forecast into a buying policy for every SKU, balancing service level against cash.
Adaptive safety stock
Buffers sized to real demand variability and your target service level, not a flat guess.
Smart reorder points
Knows when to reorder and how much, month by month, from the live projection.
Lead-time and vendor aware
Factors each vendor's lead time, MOQ, and cost to pick the right source automatically.
Per-SKU or company defaults
Set policy once at the company level, override per item where it matters.
CAPABILITY 05 · SIMULATION
Simulate before you commit.
Change an assumption and watch inventory and cash respond before a single PO goes out.
Scenario A: Order nowBASELINE
Service level ●●●●○Cash committed: Higher
Scenario B: Split the buyRECOMMENDED
Service level ●●●●●Cash committed: Lower
Capability 06 · Action Cards

Do it,
or automate it.

Predictiv hands planners the specific move. Take it yourself, or let an agent trigger it within your guardrails. Human-in-the-loop or fully automated. You set the rules.

YOU ACT
Review and execute in one click.
Every card carries the projection behind it, the recommended quantity, and the deadline. Turn it into a PO or plan without leaving the screen.
AGENT ACTS
Fires automatically inside your rules.
Set the guardrails once: thresholds, categories, spend caps. The agent handles the routine moves and logs every action for review.
The Agent · AI-Assisted S&IOP

Revenue Execution
& Planning Agent.

Sales and revenue leaders ask a what-if. The agent runs it through Predictiv's models and returns the full impact, commercial, financial, and operational, in one place.

REVENUE LEADER ASKS
"Pull the holiday promo two weeks earlier and target +15% units on the top category. Can we, and what does it cost?"
Commercial
Revenue target reachable
Units up ~15% in category
Two SKUs are the swing
Financial
Extra pre-buy spend, modest
Cash out ~3 weeks earlier
Margin holds after promo
Operational
3 SKUs short at current plan
Pre-buys timed to lead times
One vendor split to de-risk
AGENT RECOMMENDATION
Approve the earlier promo, draft the three pre-buy POs, and split the SKU-3310 order. One click to execute.
One question. One connected plan across sales, finance, and operations.
Capability 07 · Open Access

Your models and data,
inside any AI.

Predictiv ships an MCP server that exposes our proprietary models as tools. Your team asks in plain language from Claude, or any AI assistant they already use, and gets governed, company-scoped answers.

ASK IT ANYTHING
"Which SKUs will stock out this month?"
"Explain the safety stock on SKU-2048."
"Simulate a two-week supplier delay."
1
Governed & company-scoped
Every answer respects your permissions and data boundaries. No leakage across accounts.
2
Self-serve, no report queue
Analysts stop waiting on the data team. They ask, they get forecasts, projections, and POs.
3
Open standard, any MCP client
Claude, or whatever your team already uses. We meet them where they work.
Integrations

Connected to everything
that touches your inventory.

Pre-built connectors for the systems commerce brands run, plus flat-file import for everything else. If your system has an API, we can probably read it. No custom engineering, no rip-and-replace.

Shopify
Storefront
Amazon Seller
Marketplace
Amazon FBA
Fulfillment
BigCommerce
Storefront
WooCommerce
Storefront
NetSuite
ERP
Cin7
Inventory / OMS
Brightpearl
Retail OS
ShipBob
3PL
Linnworks
Inventory / OMS
Walmart
Marketplace
TikTok Shop
Social Commerce
Faire
Wholesale
NuORDER
Wholesale
JOOR
Wholesale
Recharge
Subscriptions
Skio
Subscriptions
Klaviyo
Marketing
QuickBooks
Finance
Google Ads
Demand Driver
DATA INGESTION
Don't see your system? You're probably still covered.
Custom ERP, regional 3PL, homegrown WMS. If it has an API or exports a file, we can bring it in. However your data lives today, we'll meet it there. No services engagement, no data team required.
API
Direct connection to your systems
Flat file
CSV or Excel import for anything else
Zero
Engineering hours on your side
The Payoff

We'd rather show you the number
than claim one.

Every vendor in this category quotes an uplift percentage from someone else's business. Yours will be different. It depends on your lead times, your margin structure, and how much cash is sitting in slow movers. So here's what we do instead.

THE WORKING SESSION
Bring a slice of your catalog. Leave with your real exposure, quantified.
We load your SKUs, sales history, and lead times, then run them through the models live. In one session you'll see which SKUs are genuinely at risk in the next 90 days, how much cash your current buying policy has tied up, and what the action cards would have told you last quarter.
Your stockout exposure
Which SKUs are at risk over the real lead-time window, not a rule-of-thumb buffer.
Your trapped cash
How much working capital your current policy holds beyond what the demand distribution justifies.
Your backtest
What our forecasts would have said about last quarter, measured against what actually happened.
SERVICE LEVEL
The target you set is the one you get
Coverage calibrated from your own data, so a 95% target means roughly 5% real risk, including on your long-lead-time items.
WORKING CAPITAL
Buffers sized to variability, not fear
Safety stock set against real demand variability instead of a flat multiplier. Cash released without adding stockout risk.
PLANNING CYCLE
S&IOP in a session, not a month
Ask the RE&P agent a what-if and get the commercial, financial, and operational answer in one pass.
Why Predictiv

The moat is proprietary models
you can talk to.

AI-native, not bolted on
Models, action cards, and agents are one system. Insight and action never disconnect.
Proprietary forecasting
Purpose-built models for real demand, not a generic BI tool with a chart on top.
Sales, finance & ops in one plan
The RE&P agent ties revenue what-ifs to their financial and operational impact.
Open access
Reach it through Claude or any MCP client. Meet your team where they work.
Who It's For

Built for the
whole planning team.

Sales & Revenue Leaders
Run revenue what-ifs and see the financial and operational impact instantly.
Demand & Supply Planners
Trade spreadsheets for a live plan and a prioritized action list.
Inventory Managers
Right-size buffers and reorder points across the whole catalog.
Finance & Ops Execs
See service level and working-capital trade-offs before cash is committed.
Analysts & Data Teams
Query forecasts and recommendations directly through their AI of choice.
Procurement
Turn recommendations into vendor-aware purchase orders in a click.
Getting Started

Live on your data,
fast.

01
Connect
Bring in SKUs, sales, inventory, vendors, and open POs, however your data lives today.
02
Calibrate
Models learn your demand; you set service-level and safety-stock policy.
03
Review
Planners validate the first forecasts, projections, and action cards.
04
Run the Loop
Act on cards, simulate what-ifs, and let agents run S&IOP in minutes.
See It On Your Data

Demand intelligence
that drives action.

We'll load a slice of your catalog and show live forecasts, projections, action cards, and a revenue what-if, in a single working session.

Book a demo See how it works
Pricing scales with catalog size and channel count. We'll size it on the call. No implementation fee, no annual lock-in required.