ketteQ vs Pluto7 (2026)

Agents deploy on top of SAP IBP, Oracle Cloud Planning, Kinaxis, Blue Yonder, o9, Anaplan, Logility and homegrown systems, built natively on Salesforce and AWS, and connect to CRM, ERP, MRP and WMS. Prebuilt connectors cover SAP, Salesforce, Oracle and Microsoft Dynamics, and standard REST and OData APIs handle other systems. Salesforce is supported but not required. Inside Salesforce, the platform answers available to promise questions in under 15 seconds while a sales rep is on a call.

SourceketteQ, ketteq.com, /releases/agents and /faqsread 2026-10-05

Named integrations cover SAP ECC, S/4HANA and IBP, Salesforce, Oracle EBS, Oracle NetSuite, Informatica, Google AdTech and Excel, and each has its own integration page. Data flows into the Master Ledger through SAP connectors and Cortex enabled ingestion, from structured and unstructured sources including PDFs and emails. The platform also writes back. It deploys strategies automatically across ERP, TMS and CRM and moves from planning to execution in the connected systems instead of stopping at recommendation.

SourcePluto7, pluto7.com SAP platform post 2025-06-16, CPG post 2025-06-04 and the integrations indexread 2026-09-08

A growing roster of coordinated Intelligent Digital Agents aligns decision making across demand, supply, inventory and customer commitments, all sharing the PolymatiQ intelligence layer. The current agents are Demand, Supply, Inventory and CRM, grouped as planning agents and execution agents that coordinate from supplier to customer. The CRM Agent now carries the customer commitment work, along with promotions, fulfillment and economics. Each agent weighs thousands of possible actions and runs thousands of probabilistic experiments at once.

SourceketteQ, ketteq.com/releases/agents and ketteq.comread 2026-10-05

Pi Agent is a central orchestrator under which specialized agents work in harmony on distinct planning domains. It breaks a problem into fragments and delegates each to the agent that owns it, then automates workflows end to end and moves from planning to execution across enterprise systems.

The agents are built on Google's Agent Development Kit and communicate over A2A. Pluto7 describes the result as a network of specialized agents forming an orchestrated decision layer, not a chatbot or a single model. It also says planning moves to execution in sixty seconds, but it names no mechanism behind that figure.

SourcePluto7, pluto7.com scenario planning post 2025-06-03, enterprise planning post 2025-05-28 and business value framework 2026-01-27read 2026-09-08

Agents monitor conditions in real time and continuously, reassessing scenarios and adjusting plans as demand, supply or constraints shift, with no batch runs or waiting cycles. They sense change and recalibrate as it happens, and ketteQ says a supply plan that once took 4 to 6 hours now runs in 25 seconds.

SourceketteQ, ketteq.com/releases/agents and ketteq.comread 2026-10-05

Agents act on real time signals instead of waiting for a user request. They sense real time external signals (promotions, weather, market trends, demand shifts and surges) and act on them, and real time demand sensing is a named capability. Pluto7 describes a shift from periodic manual planning cycles to always on adaptive intelligence, in which decisions are elevated before the planner asks the question. Results go to the platform and the Agentspace surface. Pluto7 names no webhook, schedule or external channel.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and enterprise planning post 2025-05-28read 2026-09-08

PolymatiQ grounds decisions in the enterprise's live planning data across ERP, MRP and WMS, combining probabilistic modeling with high speed multi scenario analysis, not static assumptions. Its solvers evaluate thousands of scenarios at once to show variability, risk and tradeoffs, and planners can change parameters, assumptions and scenarios in real time.

SourceketteQ, ketteq.com/platformread 2026-10-05

The Master Ledger is a maintained single table view built on Google BigQuery. It unifies structured ERP data with unstructured sources such as PDFs and emails into one governed, real time dataset spanning inventory, production, planning, fulfillment, finance and external signals. The ledger persists beyond any single run, every agent reads from it, and Pi.Semantic is the semantic layer over it. Pluto7 describes no retrieval or index mechanism beyond naming that semantic layer.

SourcePluto7, pluto7.com ride-share post 2025-08-11, SAP platform post 2025-06-16 and business value framework 2026-01-27read 2026-09-08

There is no stated memory store for the agents, for each agent or for each user, and no learning that changes what the agents are allowed to do. PolymatiQ learns from outcomes and tunes itself over time, and that learning lives in the solver, not in memory the agents keep.

SourceketteQ, ketteq.com/platform and /releases/agents; ketteq.comread 2026-10-05

Pluto7 describes no agent state with a stated scope and lifetime and no separate memory store for each agent. The shared Master Ledger and the persistent digital twin give a unified real time view of enterprise data on BigQuery. Together they form an application data layer of customer data that all agents read from.

SourcePluto7, pluto7.com business value framework post 2026-01-27read 2026-09-08

Agents operate independently but never blindly, executing within policies, objectives and guardrails defined by the business, with humans focused on strategy, oversight and intent. Guardrails can include business rules and approval workflows, and every decision is open to human override. ketteQ calls this agent led, human guided planning.

SourceketteQ, ketteq.com/releases/agents and /faqsread 2026-10-05

Under its Glassbox methodology, Pluto7 says AI augments planners instead of replacing them, preserving accountability, and planners become curators of intelligent decisions. Pluto7 names no approval gate, review and approve surface, autonomy setting or escalation step between an agent recommendation and its execution. The platform instead moves from planning to execution across enterprise systems in under sixty seconds and deploys strategies automatically into ERP, TMS and CRM.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and pi-agent pageread 2026-09-08

ketteQ reports SOC 2 Type II compliance, with AES-256 encryption at rest, TLS 1.2 or higher in transit and strict tenant isolation. An enterprise security framework covers end-to-end encryption, granular access controls and continuous monitoring on Salesforce and AWS. ketteQ publishes no ISO 27001 certificate and links no trust page.

SourceketteQ, ketteq.com/faqs and /platformread 2026-10-05

Pi Shield is a managed security practice built on Google's tooling (Cloud DLP, KMS, Confidential Computing, Secret Manager, Cloud VPN, IAM, VPC Service Controls, Cloud Identity, Identity-Aware Proxy and Cloud Armor). Pluto7 sells it alongside SOC services, managed detection and response and DevSecOps engagements, and calls it the intelligence layer that sits on top of the cloud, not a replacement for it. The Google Cloud certifications belong to Google as the host, not to Pluto7, and the delivery services are engagements, not controls built into the platform.

The product runs inside the customer's own Google Cloud tenant, so the customer operates IAM and Cloud Identity over it. Agent identities are defined with the same rigor as human identities. Pluto7 names no SOC 2, ISO 27001 or other attestation of its own.

SourcePluto7, pluto7.com/security-and-governanceread 2026-09-08

PolymatiQ enables explainable, governed and coordinated decision making with immutable audit trails, and presents plans tied to a business goal and confidence score. Every recommendation keeps a complete audit trail, and each decision comes with its rationale and tradeoffs.

SourceketteQ, ketteq.com/platform and /releases/agentsread 2026-10-05

Under the Glassbox methodology, planners can see why a recommendation was made, not just the outcome, alongside data backed ROI metrics tied to the business value framework. Pluto7 names no run trace, decision log, retention period or way to reconstruct what an agent did after the fact. It speaks of "continuous evaluation of decisions" as a property of the architecture but attaches no mechanism to it. The auditing Pluto7 names under security and governance is Google Cloud compliance auditing of the tenant, not a record of agent activity.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and /security-and-governanceread 2026-09-08

Hosting is cloud SaaS on Salesforce and AWS, and ketteQ publishes no residency claim, region list, region choice or customer environment option. Its flexible deployment means agents either sit on top of an existing planning system or run on ketteQ's own platform, and both run as SaaS. The first use case goes live in four to eight weeks above existing systems, with no data migration, and tenants are strictly isolated.

SourceketteQ, ketteq.com/faqs and /releases/agents; ketteq.comread 2026-10-05

Under the Glassbox methodology, customers own their data and models, and all intelligence runs inside the customer's own Google Cloud environment. The platform is delivered into that tenant through a Google Cloud Marketplace private offer. Because the platform runs in the customer's own cloud tenant, the customer chooses the project, region and controls instead of picking from a list Pluto7 sets. Pluto7 publishes no region list. Deployment is Google Cloud only, and Pluto7 names no AWS, Azure or on premises option.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and the Cloud Marketplace listingread 2026-09-08

ketteQ ships a growing roster of coordinated prebuilt Intelligent Digital Agents, each focused on a specific domain, including Demand Planning, Inventory and Customer Commitment agents. The four agents named today are Demand, Supply, Inventory and CRM. The platform's modules span planning (IBP, supply, demand, inventory and service parts) and execution (control tower, fulfillment and allocation, asset management, work orders, and service and part delivery).

SourceketteQ, ketteq.com/releases/agents and /platformread 2026-10-05

Three named agents ship. A Demand Agent handles forecast accuracy, an Inventory Agent handles positioning and safety stock, and a Finance Agent handles margin and cost to serve tradeoffs, and they also go by Ron, Kassy and Alex. Operations in a Box covers finance, marketing, HR, IT and sales. Each of the three agents is a separate unit a customer could adopt on its own. Pluto7 has no catalog, gallery or template library page where customers browse and select among agents.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and CPG post 2025-06-04read 2026-09-08

PolymatiQ is a proprietary, patent pending agentic AI solver built for supply chain planning, and there is no stated customer choice of LLM or model routing. The solvers power Quintus, ketteQ's planning product, and return an answer in seconds.

SourceketteQ, ketteq.com/platform and /faqsread 2026-10-05

The model stack is Gemini 2.0, Vertex AI and the Cortex framework on Google Cloud, with agents built on Google's Agent Development Kit. Google is the single provider. Pluto7 describes no way for customers or admins to select a model, no routing control and no bring your own model option. When the Glassbox methodology says customers own their models, it means the trained artifacts inside the customer's tenant, not a choice of which foundation model runs.

SourcePluto7, pluto7.com scenario planning post 2025-06-03 and business value framework 2026-01-27read 2026-09-08

As a Salesforce-native application, ketteQ's objects are exposed through the Salesforce platform's API layer, which customers can use to extend it. ketteQ also names standard REST and OData APIs, but publishes no developer portal, API reference, SDK or MCP server of its own.

SourceketteQ, ketteq.com/releases/agents and /platform; ketteq.comread 2026-10-05

The agents are built with Google's Agent Development Kit and, Pluto7 says, follow the A2A and MCP protocols. Pi Agent is published on Google Agentspace and Gemini Enterprise and listed on Google Cloud Marketplace, so outside callers can in principle reach the agents through Google's agent surfaces. Pluto7 publishes no A2A agent card naming an endpoint and an auth scheme. It names no endpoint, scope or reference behind the MCP claim, and it has no API, SDK or developer portal of its own. There is no docs or developer subdomain, only a Zoho hosted help center behind sign in.

SourcePluto7, pluto7.com scenario planning post 2025-06-03 and Gemini Enterprise page, probe recordedread 2026-09-08

PolymatiQ's thousands of multi-pass experiments keep tuning the engine's own assumptions against realized outcomes after deployment, and its outputs are plans tied to a business goal with a confidence score. The customer's plans feed that loop, and what it improves and measures is the solver. Customers report results such as 4 times better demand planning accuracy at Trimble Transportation and $17 million in inventory savings at quip.

SourceketteQ, ketteq.com and /platformread 2026-10-05

Before execution, thousands of digital twin simulations run to find the optimal path. They test the customer's supply chain plan (alternates, lead times, tariff shifts, demand scenarios), which is the product's own function, not the agents. Pluto7 names no scored result, comparison, holdout or post deployment tuning loop that measures agent decisions, and no configuration or readiness gate. It presents "continuous evaluation of decisions" as a property of the Master Ledger architecture, with no mechanism attached.

SourcePluto7, pluto7.com business value framework post 2026-01-27 and pi-agent pageread 2026-09-08

There is no stated browser or computer use. Agents work through planning system integrations and carry out decisions there. Those connections run through prebuilt connectors and standard REST and OData APIs.

SourceketteQ, ketteq.com/releases/agentsread 2026-10-05

The agents reach the customer's systems through SAP connectors, Cortex ingestion and write back into ERP, TMS and CRM, and they act inside the platform and the Agentspace surface. Pluto7 describes no browser, desktop or remote control of an interface it does not own.

SourcePluto7, pluto7.com SAP platform post 2025-06-16 and pi-agent pagesread 2026-09-08