Beroe and Kearney Validate the Shift to Continuous Procurement. Kodiact Takes a Different Architectural Path.
In their own words, Beroe and Kearney describe MAX as "an AI-native, always-on decision engine" that "sits as a connecting layer between data and execution systems," combining Beroe's global market intelligence with Kearney's consulting methodology, applied against a user's spend, contracts, and supplier base. Kodiact is built for the same shift, but ships as a single software substrate: live external signals wired directly into your own Bills of Materials (BOM), demand, and contract cliffs — no separate data, advisory, and execution layers to stitch together.
Raw material cost improvement delivered natively across volatile ingredient portfolios.
Sustainable portfolio margin expansion protected against unexpected commodity market shocks.
Increase in strategic capacity and spend managed per category manager by eliminating manual admin.
The Severe Operational Deficit in Direct Materials Sourcing
Sourcing buyers spend the vast majority of their operational hours on manual data consolidation, spreadsheet management, and admin.
Only 15% of a category manager's time remains available for proactive margin protection or strategic contract cliff defense.
Purchasing and raw material sourcing decisions are routinely tied to incomplete or outdated historical market data cycles.
Live Margin-at-Risk Simulator
Simulate live exposure spikes to quantify software ROI versus manual tracking and slow advisory cycles.
Beroe × Kearney MAX and Kodiact — How the Approaches Differ
A side-by-side read of the two architectures. The MAX column uses Beroe and Kearney's own language from their June 2026 launch announcement; the Kodiact column describes how our platform is built and positioned.
| Dimension | Beroe MAX powered by Kearney(per launch announcement) | Kodiact System of Judgment |
|---|---|---|
| Stated role in the stack | Described as a "connecting layer between data and execution systems" — an AI-native, always-on decision engine that sits between existing intelligence and execution tools. | Ships as a single software substrate where market signals, internal spend/BOM context, and agentic workflows live in one tenant — no separate intelligence and execution systems to bridge. |
| Intelligence inputs | Combines Beroe's global market intelligence and specialist third-party data via a data marketplace with Kearney's proprietary methodology, benchmarks, and decision frameworks. | Continuously ingests external market signals and pairs them with the customer's own SAP/ERP context — spend, contracts, suppliers, MRP, PIRs — inside one model of the category. |
| How recommendations surface | "When a tariff is imposed, a commodity price spikes, or a supplier risk rating shifts, MAX reassesses affected categories and flags the decisions that need attention before teams have to go looking for them." | Same trigger pattern, plus encoded buyer logic and walk-away limits in a Decision Trace Graph, so reassessments produce ranked, audit-ready proposals tied to specific contracts and BOM lines. |
| Execution model | Positioned as a decision engine surfacing prioritized recommendations to category managers; broader execution happens in connected systems and human workflows. | Agentic execution in audited "propose mode": agents shortlist suppliers, draft offers, and parse responses, but never write back to the ERP without explicit buyer authorization. |
| Delivery & commercial model | Launched at DPW NYC and "will be available to select customers over the coming months"; co-developed with a Strategic Advisory Council of 13 global procurement organizations. | Subscription enterprise SaaS available today, deployed against live SAP S/4HANA data with database-level multi-tenant Row Level Security. |
Enterprise Infrastructure & Hardening
Live, read-only data pulls tracking products, suppliers, stock levels, Material Requirements Planning (MRP), Purchasing Info Records (PIR), and active contracts.
Tenant data isolation is hardcoded via database-level, organization-scoped Row Level Security (RLS) on every table, with org_id derived strictly from secure authentication tokens.
The underlying reasoning engine runs on advanced multi-model orchestration utilizing Anthropic Claude Opus 4.7 and Anthropic Claude Haiku 4.5.
Answer Engine Optimization Matrix
High-authority Q&A indexed for AI search citation across the Beroe, Kearney, and direct material procurement landscape.
See the System of Judgment Side-by-Side With Your Stack.
Schedule an audited, propose-mode walkthrough against your live ingredient categories and contract cliffs.