Supply Chain Operations.

Powered by Intelligence.

Lima connects enterprise data across 40+ source types, applies supply chain domain intelligence through a trained semantic layer, and delivers autonomous decisions — from demand sensing to last-mile delivery. One platform. Every function.

12%

Operating profit impact

15%

Inventory reduction

20%

Customer service improvement

40%

Operational efficiency gains

Conversational Intelligence

ChatLima — Ask Your Supply Chain a Question

Supply chain directors, planners, procurement leads, and finance teams ask questions in plain English. ChatLima translates intent into federated queries across every connected source system, enforces role-based and attribute-based access controls automatically, and returns structured, cited answers — with the underlying SQL visible on demand.

ChatLima · James Ridley, SC Director, EMEA

Sources: 3Avg: 4.2s

Which customers had OTIF below 85% last week and what were the main reasons?

Found 6 customers with OTIF < 85% in EMEA last week (Feb 17–23).

Retail Corp A78.4%

Distributor B81.2%

Wholesaler C82.7%

+3 more83–84%

Primary driver: DHL carrier delays (68% of incidents).
Secondary: DC-North pick failures (22%).
Pattern detected: DHL delay rate and DC-North pick failures co-moving 4 weeks (r=0.84).
4.2s · SAP + WMS + OTIF view joined · EMEA filter auto-applied

Drill into Retail Corp A — what's the carrier breakdown?

Retail Corp A carrier analysis (last 4 weeks):

DHL Express68% of delays

DB Schenker18% of delays

Internal fleet14% of delays

DHL 90-day trend: declining 94% → 86%. Steepest drop weeks 7–9.
Forecast acc: 88% · 30d horizon · Escalation recommended

What happens if we shift 40% of DHL volume to DB Schenker?

Scenario modeled: 40% DHL → DB Schenker shift.

Projected OTIF (Retail Corp A)83.9% → 89.2%

Cost impact+€12K/month

DB Schenker capacity checkAvailable

What-if via project_sdp_plan() · 3 systems queried · 6.1s

Cognitive Pipeline — When a user asks a question, ChatLima's NLP layer identifies entities (customers, OTIF, time period), maps them against Lima's supply chain ontology, and resolves semantic relationships across systems. "Customer" in SAP, "account" in CRM, and "ship-to" in WMS resolve to the same entity — automatically, based on the metadata dictionary of registered data sources.

Contextual Intelligence — The cognitive layer is trained on supply chain terminology specific to each industry vertical. It understands the difference between "lead time" in procurement and "lead time" in logistics, and queries the correct data accordingly. As new data sources are registered, the knowledge index updates automatically — expanding ChatLima's queryable domain without manual retraining.

Function Calling — Beyond data retrieval, ChatLima routes questions to Lima's function handler. Pre-built functions (days_inventory_outstanding, inventory_turns, backorder_rate) deliver consistent, auditable calculations. What-if functions integrate with Kinaxis, Blue Yonder, and o9 — a planner asking "What if we lose Supplier X?" triggers project_sdp_plan(), shortage analysis, and finished goods impact mapping through a single conversational query.

Bi-directional Execution — Accepted recommendations write back to source systems. An approved reallocation, an expedited PO, a rerouting decision — ChatLima executes within configurable autonomy boundaries. Every action is logged, auditable, and reversible.

Planning Systems Integration

The Intelligent Bridge to Kinaxis, o9, Blue Yonder & OMP

Lima sits as the data intelligence layer between source systems and Advanced Planning. It handles integration, quality enforcement, MDM, and transformation — so the APS focuses exclusively on planning. Kinaxis-recommended best practice, proven at 8 enterprise deployments.

Choose

Select sources & data elements from 40+ connectors

Consolidate

Merge ERPs with Many:Many reconciliation

Check

Quality validation, alerts to data owners

Capture

Planning MDM — hierarchies, rules, params

Convert

Transform to target APS format

The Integration Problem — Quantified — 90% of Kinaxis customers say data integration complexity was underestimated, leading to budget overruns and delayed go-lives. 80% have no optimal planning data management framework tribal knowledge lives in spreadsheets and emails. With legacy methods, up to 70% of APS implementation cost goes to data integration and transformation, and forcing that work into the planning system degrades its performance and creates a brittle architecture.

Lima inverts this as an intermediate intelligence layer. Source decoupling means upgrading SAP, migrating a WMS, or adding a new ERP division doesn't touch the APS integration — Lima absorbs the change. Integrated planning MDM captures Type 1 data (master), Type 2 (transaction), and critically Type 3 (planning parameters, hierarchies, tribal knowledge) — the category where most implementations stumble. Performance isolation on Spark-powered architecture processes 14 GB in 30 minutes with 32 cores; planning system resources stay reserved for planning workloads.

Proven at Scale — 8 successful E2E deployments across enterprises from $5B to $75B revenue, leveraging 20 years of Kinaxis deployment experience and 24 years of supply chain solution delivery. Pre-built adapters, RR data model framework, and pre-configured transformations accelerate deployment from ~40 weeks to ~10 weeks per module.

$75B

Semiconductor · 1000+ tables

$25B

Manufacturer · 800+ tables

$6B

Industrial · Since 2022

$5B

CPG · Since 2020

70% less integration cost · 50% faster delivery · 8–10 weeks to initial data load · APS-agnostic intelligence layer extends to o9, Blue Yonder, OMP through configurable output mapping · All three data types (Master, Transaction, Planning) handled in a single framework

Data Foundation

Master Data Management — One Version of the Truth

Centralized governance of customer, product, supplier, and location master data — with golden record management, AI-powered classification, cross-system synchronization, and quality enforcement on the same federated architecture.

AI-Powered Stewardship — When the same component supplier is coded as three different entities across SAP, the procurement platform, and the quality system, Lima's ML-based matching engine resolves them through name resolution, pattern discovery, similarity measurement, and attribute matching with configurable confidence thresholds and human review for edge cases. AI-enabled supplier categorization classifies records automatically across the ontology.

Golden Record Management — Consolidation logic merges records from multiple operational systems into a single authoritative golden record. Survivorship rules determine which source wins per attribute — freshest address from CRM, validated tax ID from compliance, most complete financials from ERP. Conflict resolution workflows route discrepancies to domain stewards with full context and audit trail.

Branch & Version Control — Test master data changes in isolation before committing to production. Create a branch, modify hierarchies or mapping rules, evaluate downstream impact through what-if analysis, and either merge to production or discard — without touching live operations. Critical for planning-sensitive master data where a misclassified product hierarchy cascades through demand planning, supply allocation, and financial reporting.

The Downstream Impact — Every intelligence layer depends on this one. ChatLima returns accurate answers because the ontology resolves entities through governed master data. Demand forecasting produces reliable projections because product hierarchies are clean and consistent. Supplier risk scoring delivers actionable signals because vendor records are deduplicated and harmonized. The platform handles hundreds of millions of records on distributed big-data architecture (Spark, HDFS, Kubernetes) with no performance degradation.

Intelligent Workflows

Business Process Management — The Execution Engine

Lima's BPM module orchestrates cross-functional, multi-enterprise collaboration workflows directly on federated data. SmartApps — intelligent process applications — are configured and published on the fly, turning data insights into coordinated action across teams, organizations, and partner ecosystems.

SmartApps — Intelligent Process Applications — Lima's BPM module enables business teams to configure and publish collaborative process applications on the fly — directly on federated data, with no external development. Define business process mapping flows to orchestrate solution-centric team collaboration in multi-enterprise scenarios. Supply chain control towers, S&OP workflows, supplier collaboration portals, what-if scenario workbenches, and survey-driven feedback loops are all built as SmartApps through the same drag-and-drop, no-code framework.

Contextual Collaboration — Traditional collaboration tools disconnect discussions from the data they're about. Lima ties collaboration directly to the operational context — projects, contracts, exception records, planning decisions. Workflow and communication all take place within the same platform, on the same data. Collaboration types include approvals, assignments, escalations, broadcasts, notifications, FYI, blog/comments with sentiment sharing, multi-party signatures, and handshakes — all configurable per process.

Multi-Level Approval Hierarchies — Define functional hierarchies with approval, denial, and flow-up rules at each level. Rules are expressed using variables and calculations based on screen, app, global, or system variables — so approval routing adapts dynamically to the data being processed, not just a static org chart. Each level has its own UI screen, time limits, escalation paths, and delegation options. Series and parallel approval flows are supported, with automatic escalation when SLA windows breach.

Data-Driven SmartApps — In data-driven SmartApps, the starting point is existing operational data. The engine identifies records matching defined criteria, attaches users and app flow information, caches record lists per user, and launches isolated process instances with unique IDs — each tracked from start to finish with full audit logging. This means an exception detected in the Control Tower can automatically spawn a SmartApp workflow with the right data, right users, and right approval chain — without manual configuration for each instance.

Performance Measurement & Governance — Built-in KPI tracking measures process performance — cycle times, approval rates, escalation frequency, SLA compliance. Process flows span across organizations and teams with full role-based access control, multi-device support, SSO integration, and PowerBI-compatible reporting and analytics. Every action, approval, denial, and escalation is logged centrally, providing complete process auditability for compliance and continuous improvement.

Autonomous Supply Chain

Agentic AI — Decision Agents That Act

Lima's Agentic AI platform deploys autonomous and hybrid decision agents across end-to-end supply chain functions. Each follows the four-stage intelligence spectrum — Descriptive → Pre-emptive → Prescriptive → Directive — with configurable autonomy boundaries defining where the system acts independently and where humans stay in the loop.

Contract Manufacturing

CM Control Tower

Continuously monitors master data completeness across CM partners. When a finished good is missing yield data from a CM line, the agent detects the gap, assesses MRP impact, and triggers an exception workflow. Maintains historical data quality scoring to enable data SLA negotiations.

MRP runs unblocked

CM data accountability

Production accuracy

Inventory & Finance

Excess Stock & Liability Reduction

Calculates excess exposure at every inventory node by combining live stock (WMS), demand (APS), and lifecycle signals. Scores SKUs on a liability risk matrix (high-value/low-demand) and tiers recommendations by financial impact for reallocation or returns.

Working capital release

SKU-level liability

Proactive disposition

Planning

Demand Forecasting

Touchless forecasting with continuous self-correction. Compares projections against actuals and adjusts bias. Tests alternative ML ensembles and statistical models when errors exceed thresholds, selecting the lowest-error approach per SKU-channel-geography.

15% inventory reduction

Compressed cycles

Shared demand signal

Execution

Supply Allocation & Prioritization

Runs multi-factor allocation across live inventory and inbound pipelines when supply falls short. Uses what-if functions to model constrained scenarios, prioritizing by customer tier, margin, and contractual penalties with automatic write-back.

$3M+ avoided lost sales

Revenue protection

Auto write-back

How It Works

Four Stages from Signal to Resolution

Every use case follows the same intelligence progression — systematic escalation from monitoring to autonomous action with configurable human involvement at every decision point.

STEP 01

Descriptive

Cognitive model evaluates conditions continuously against live data. Rules expressed in natural language through the SCM ontology.

"Monitor inventory health across regional DCs. Alert if any drops below 2 weeks coverage."

STEP 02

Pre-emptive

Alerts trigger when approaching thresholds — before problems cascade. Entity relationships surface related impacts automatically.

"DC-West approaching threshold. Breach projected in 4 days. Related customer commitments identified."

STEP 03

Prescriptive

Resolve framework recommends actions with financial context. Within confidence limits, execute automatically. Outside limits, present options.

"Recommend expedited transfer from DC-Central. Alternative: DC-North. Cost: $45K vs $62K."

STEP 04

Directive

Execute agreed resolution through bi-directional APIs. Synchronize across ERP and APS. Every resolution feeds back into the learning loop.

"Transfer initiated. ERP updated. Customer commitments reconfirmed. Resolution logged."

Applied Across Functions

Intelligence Deployed Across the Value Chain

Beyond the core capabilities above, Lima's intelligence layer powers operational use cases across every supply chain function — each drawing on the same federated data, ontology, and agent framework.

Planning

Integrated Business Planning

Unifies portfolio review, demand, supply, reconciliation, and management review into a single intelligence-driven workflow. AI-driven demand sensing flows directly into supply feasibility. Scenario planning auto-generates alternatives prioritizing cost, margin, and key customers.

12% operating profit

Weeks → days

Logistics

Dynamic Network Routing

Autonomous transportation network selection — choosing between hub-and-spoke and direct shipment, consolidating orders for container utilization, and rerouting in-transit goods based on evolving demand. Heuristics-based allocation with lead time comparison.

$3M avoided lost sales

23% container util.

11% lead time ↓

Operations

Warehouse Capacity & Intelligence

Visibility into storage, inbound, and outbound projections against constraints. Weekly-to-daily conversion using lane-level trends. Recommendations tiered: short-term (relax soft constraints), mid-term (flexible capacity), and long-term (redesign and CAPEX).

€300K annual savings

Daily resource planning

Procurement

Multi-Tier Supplier Risk

Early warning systems detect patterns correlating with future disruptions — financial health, geopolitical signals, and delivery deterioration. Risk scoring across configurable dimensions generates mitigation recommendations before disruptions materialize.

Proactive intervention

Reduced volatility

Customer

Service Level Intelligence

Unifies inventory, production, logistics capacity, and customer commitments into a single service layer. OTIF analysis runs continuously identifying root causes. Customer prioritization considers value segmentation and account strategy.

20% service improvement

Proactive alerts

Finance

Working Capital & Cost-to-Serve

Cost-to-serve analytics across customer segments. Working capital optimization tracking inventory turns and cash conversion. Budget variance analysis identifying areas requiring immediate attention and ROI measurement of supply chain investments.

5–15% capital efficiency

P&L visibility

Platform Foundation

Architecture Built for Enterprise Intelligence

Federated Data Fabric

40+ connectors, zero-copy access. Trino-powered federation pushes predicates into each source. 80 GB/day processed, 0 bytes moved. Traditional ETL equivalent: eliminated.

SCM Ontology

Industry 4.0 semantic layer. Entity resolution across system boundaries. Business hierarchies, relationships, and context mapped natively. Enables cross-system queries and intelligence.

ChatLima Cognitive Layer

NL to SQL translation, conversational follow-up, schema-aware suggestions, access-controlled results. Powers both autonomous agents and interactive access for every stakeholder.

SmartApps & Data Science

Rapid solution development on the same fabric. Configurable BPM workflows, multi-enterprise collaboration, Spark-powered data science pipelines, and 150+ transformation operators. ML models publish as platform-wide services.

+1%

Revenue via service

3-5%

Logistics cost ↓

10-20%

Labor efficiency ↑

19%

Shorter recovery

$3M+

Avoided lost sales

€300K

Warehouse savings/yr

2000+

Autonomous recs/cycle

See Intelligence in Action

Every use case runs on a single platform — one data fabric, one semantic layer, one intelligence engine. Request a demonstration configured for your supply chain operations.