AI finds the possibilities
AI helps each of the ten business agents find the next best move, from repricing to routing to recovery, across every marketplace you sell on.
Flagship platform by Sparande · Commerce intelligence
SparGrid: the complete commerce platform
SparGrid is the full platform behind the Sparande product suite: ten business agents and more than sixty capabilities working over one Unified Commerce Model across Amazon, Shopify, eBay, BigCommerce and WooCommerce. AI helps the agents find the next best move, from repricing to routing to recovery, and deterministic, quantitative mathematics verifies every one before it can be accepted. Each recommendation is then explainable, approval-gated, and reversible.
The focused Sparande apps each do one job. SparGrid runs the whole decision system, and you can adopt it directly.
10 business agents60 capabilitiesOne Unified Commerce ModelOperator-approved, reversible
AI you can launch with confidence
AI suggests. Deterministic maths decides. Commerce decisions are where AI can earn you the most and cost you the most, so the platform splits the work for every agent. AI does what it is good at, spotting opportunities across every channel. Mathematics does what it is good at, proving whether an action is safe. AI never sets a number, and nothing reaches a marketplace on the model's word alone.
Maths verifies what is safe
Every opportunity must name the quantitative engine behind it. Deterministic, quantitative mathematics sets every number: same inputs, same outputs, every time.
You decide what launches
Nothing is written to a marketplace until an operator approves it. Every dispatch is logged and can be rolled back.
Decisions the team can trust
Most tools stop at a chart. SparGrid produces the decision (the reprice amount, the reallocation quantity, the bundle to launch) as a repeatable output you can inspect, approve, and reverse. This is governed decision support, not blind automation.
- Unified Commerce Model across Amazon, Shopify, eBay, BigCommerce and WooCommerce
- AI to find opportunities, quantitative mathematics to verify every one
- Confidence and drivers exposed on every output
- Operator approval and a kill-switch on every dispatched action
- Time-stamped, audit-ready traceability by design
How SparGrid works
SparGrid connects to your channels through their official APIs, normalises catalogue, orders and inventory into one model, runs governed recommendation workflows over it, and emits operator-approved actions back to the marketplaces, with every step logged and reversible.
Governed pipeline
Data in
- Amazon, Shopify, eBay, BigCommerce, WooCommerce
- Catalogue, orders, inventory & fees
- Warehouse / ERP & cost feeds
- Normalised into the Unified Commerce Model
Governed recommendation workflows
- Velocity, ageing & demand-drift monitoring
- Product relationship and bundle discovery
- Stock-risk, marketplace routing & net-margin
- Agents orchestrate; the logic remains consistent and reviewable
Operator-approved actions out
- Reprice, rebalance, bundle, replenish, route
- Explainable rationale per recommendation
- Immutable audit-log entry
- Dispatched to the marketplace on approval
Reads by default; writes only after operator approval. Every dispatch is logged, reversible, and covered by a kill-switch.
In a private walkthrough, we can show live channel connections, recommendation rationale, the approval queue, and audit exports.
Ten business agents. Sixty capabilities.
SparGrid organises sixty platform capabilities into ten business agents. A focused Sparande app draws on one or two of these agents; the full platform runs all ten over the same Unified Commerce Model, the same governance, and the same audit trail. The real value is not any single capability, it is the composition of all sixty into governed decision workflows.
Revenue Growth
- Promotion opportunities
- Quantity breaks
- Post-purchase upsells
- Countdown campaigns
- Mix and match
- Bundle deals
- Campaign monitoring
Merchandising
- Bundle discovery
- Product affinity
- Cross-sell opportunities
- Upsell opportunities
- Canonical product graph consumption
- Bundle lifecycle management
Inventory Health
- Stock risk
- Slow-moving detection
- Reorder signals
- Early intervention
- Inventory scoring
- Stockout and overstock risk
- Anomaly monitoring
Inventory Recovery
- Markdown recovery
- Flash-sale recovery
- Recovery bundles
- Marketplace routing
- Wholesale routing
- Liquidation export
- Expected net recovery
- Returns recovery
Margin Protection
- Pricing optimisation
- Margin leakage
- Price conflict
- Marketplace fees
- Cost drift
- Channel profitability
- Price-floor validation
Marketplace Operations
- Channel state
- Listing state
- Cross-channel pricing
- Marketplace readiness
- Connector health
- Sync status
- Inventory rebalancing
Forecasting
- Demand forecasting
- Seasonality detection
- Trend projection
- Replenishment forecasting
- Forecast accuracy
Compliance
- MAP validation
- Brand policy
- Approval policy
- Price floor
- Inventory floor
- Hard blocks
- Audit
Attribution
- Incremental revenue attribution
Ties every approved action to the revenue it actually moved, so impact is measured rather than assumed.
Finance
- FX normalisation
- Settlement ledger
- Commission calculation
- Finance reports
- Invoice support
A focused app, or the full platform
You do not have to adopt everything at once. Start with a focused Sparande app that owns one job, or run the whole platform. Both are powered by the same SparGrid agents, governance, and audit trail, so nothing is thrown away when you grow from one into the other.
A focused Sparande app
from $14 / month · powered by SparGrid
- Owns one commercial job, drawing on one or two of the ten agents
- Embedded directly in your Shopify store
- Reports against a single, clear business outcome
- The fastest way to put SparGrid to work on a specific problem
- A growing family: InventoryIntel, Velocity Pulse, Bundle Optimizer, Margin Optimizer, with more to come
The SparGrid platform
from $150 / month · the complete system
- All ten business agents and sixty capabilities
- Every channel unified into one Unified Commerce Model, not just Shopify
- Cross-channel decisions the focused apps cannot make on their own
- The full governance, approval, and audit surface across the suite
- For teams that want one decision system for the whole operation
Operator workflow
The SparGrid UI is built around one sequence: review the recommendation, inspect the rationale, approve or dismiss the action, and then watch the result land in the action log.
Recommendation queue
Surfaces the next best action, grouped by channel and urgency so operators can review what matters first.
Rationale view
Shows the drivers behind the recommendation, the confidence level, and the expected impact before any action is approved.
Action log
Records every approval, dispatch, and rollback so the team can audit what happened and when.
Representative product views
The views below are non-sensitive representations of the SparGrid surface: health state, recommendation rationale, action logging, the approval pipeline, and onboarding.
SKU healthAt Risk
Cover 21dHealthy
Cover 6dAt Risk
WatchCritical
Rationale view3 drivers
Velocity (7d)-22%
Amazon cover6 days
Net marginHolds
Action logExportable
10:35:14Reprice proposed
10:36:02Operator approved
10:36:05Dispatched to eBay
Action pipelineOperator-gated
Channel connectors→Governed recommendation workflows→Approval queue + dispatch
Recommendations wait in an approval queue. Nothing is written to a marketplace until an operator approves it, and every dispatch can be rolled back.
OnboardingScoped path
- Connect your channels
- Sync catalogue & orders
- Shadow recommendations
- Go live with approvals
Built for accountable automation
Human oversight
SparGrid provides analysis and recommended actions. The decision to dispatch (and accountability for it) remains with your operators, under your approval workflow.
- Recommendations, not silent changes
- Clear approval and accountability boundaries
- Designed for review, challenge, and rollback
Data posture
- Official OAuth connections to each channel
- Credentials held in a managed secret vault
- Per-tenant isolation: no data pooled across merchants
AI finds the possibilities. Our maths verifies what is safe to launch.
This is the moat, and the platform enforces it for every agent. Each opportunity must name the quantitative engine behind it, and an AI proposal cannot be accepted until that engine has admitted it. AI never sets a number. That is why explainability, provenance, and reversibility are not add-ons: the maths behind each output can be inspected, approved, and operated inside your commercial and finance governance.
- AI proposals admitted only by quantitative models, which set every number
- Traceability per recommendation and dispatched action
- Rationale designed for operations and finance review
- Deterministic and repeatable: same inputs, same outputs, every time
Clear boundaries
SparGrid is a commerce intelligence and optimisation layer across your channels, not a replacement for your marketplaces or your team's judgement. The boundary is deliberate.
Out of scope
- Silent, unapproved changes to price or stock
- Black-box scoring with no rationale
- Replacing your marketplace or storefront
Non-negotiables
- No data pooled across merchants
- No writes without operator approval
- No automation without an audit trail and kill-switch
Commercial model
SparGrid is a subscription platform. Plans scale with the channels you connect and the volume you run. Every plan includes the full platform capabilities and agent workflows, with no per-recommendation fees.
Getting started
Connect your channels, sync your catalogue, and run SparGrid in shadow mode to see the recommendations before anything goes live. When you're ready, switch on approvals and start dispatching.