Retail and CPG run on margin. Their AI agents should know how.
The Operational Context Graph for retail, CPG, and consumer ops — merchandising, supply, store ops, fulfillment, returns. Agents that cite the benchmark. Healthcare ships today. Retail ships Q4 2026. The design-partner cohort is open — shaping the ontology before GA.
Every Amandil OCG carries the same dimensions — public-data spine, reference standards, regulatory frameworks, canonical operational example. The substance is per-vertical. Healthcare is live today; retail ships Q4 2026.
The same two sides as the live healthcare graph — the knowledge an agent reasons from, and the guardrails that govern what it does — instantiated for merchandising, supply, store ops, fulfillment, and returns. Final scope shaped with the design-partner cohort.
L1 domains across retail, CPG, and consumer ops. Reference processes with owners, metrics, and dependencies. Anchored to GS1, EDI 850/856, ISO 8000, UNSPSC. Stitched to Census, BLS, NRF, GS1.
CPSC, FDA, FTC, state UDAP. Trade-promotion clauses, audit checkpoints, recall procedures. Industry benchmark frameworks. Trace coverage on every agent action.
Three example archetypes shown on the healthcare page — Reconstructor, Pattern Analyzer, Submission Compiler — illustrate shapes the substrate produces. Each will be tuned to the retail substrate with the cohort.
Rebuilds a returns case from SKU history, channel-specific policy, prior-case patterns, and the consumer-protection clauses each category maps to.
Classifies markdown signals against category-level elasticity, peer-velocity patterns, channel mix, and the merchandising rulebook the chain operates by.
Assembles a GS1-ready data submission from the retailer's product master, cited against every attribute's source schema and trading-partner requirement.
The same three customer segments that anchor every Amandil vertical. Operators run the stores. Advisors compress discovery into branded analysis. Build partners embed the graph into their products.
"Where should we invest in AI?" You run the operation — merchandising, supply, store ops, fulfillment, returns. Most teams pick what to automate based on a vendor pitch or what feels problematic to an executive. You need a benchmark-anchored answer — and agents that cite the standard, name the SKU-level process, and return the audit trail.
You get: diagnose against the retail benchmark library → prioritize by automation score reweighted by your gaps → simulate the top candidates → ship agents grounded in the graph via MCP.
"How do we scale our retail AI engagements?" Every project opens with a discovery phase. Weeks rebuilding the same markdown workflow, the same returns flow, the same supply-chain map. The sameness goes unclaimed — and the artifact ages on contact with reality.
You get: the discovery phase compressed to a 30-minute branded analysis. White-label per client. Branded blueprints, business cases, diagnostics. Multi-client workspaces with scoped grants.
"What grounds our retail AI features?" You're shipping AI inside your product. Every LLM-integrated feature hits the same wall — the model has no grounded knowledge of how a retailer, CPG, or marketplace operates. Building the graph yourself is a multi-year, multi-million-dollar undertaking.
You get: a native MCP server. OAuth 2.0 plus headless bearer tokens, instantly revocable. Grant-scoped per partner. Versioned, validated graph surface. Multi-year build, solved on day one.
Top-100 retailers and CPGs. Design partners shape the ontology, get the substrate at cost during the cohort window, and lock pricing for two years post-GA. We don't take partners we can't deeply serve.
Tell us where you sit — where the AI-vs-ops gap is sharpest, what regulation pins your team down, and what you'd want the first agent to do. We'll come back with a roadmap shaped to that.