Back to vendors
L

Lumi AI

Visit site
Data analyst agentindependentVerified 2026-07-01

Lumi AI is a chat based multi agent data analyst that runs inside a company's network, finding supply, pricing, and procurement opportunities across enterprise resource planning systems and databases.

Lumi AI is a chat based data analyst built for the operational core of large supply, distribution, and retail businesses. A user asks a question in plain language, such as where pricing leaves margin on the table or which suppliers to consolidate, and Lumi runs the analysis and explains the answer. The company is careful to say this is not simple text to SQL conversion. Lumi uses multiple coordinated agents and a structured knowledge base to interpret intent, apply the right business definitions, run the correct analysis, and return a decision ready result, positioning itself as a new category of multi agent analytics application rather than a query translator.

The defining architectural choice is that Lumi runs inside the customer's own environment and never moves or stores data externally. All processing happens within the client network and follows the client's existing security controls, and Lumi supports both cloud and on premises deployment. It connects to enterprise resource planning systems such as SAP, Oracle, and Microsoft Dynamics, along with warehouses and databases, through read only data gateway connectors that use governed credentials and encrypted connections. Security includes role based access control, query cost and duration limits, and SOC 2 compliance, which together let regulated enterprises hand a chat analyst live access to sensitive operational data.

Accuracy comes from a knowledge base that stores field definitions, table relationships, business rules, key metrics, and aliases, so that answers reflect the true meaning of a company's own terms rather than generic assumptions. Lumi is transparent about its work, exposing the assumptions it made and a step by step explanation of how each number was produced, and it rebuilds the full analytical pipeline on each run so results stay consistent as a question is refined. For analytics teams, a guided environment helps draft and troubleshoot SQL and Python while Lumi pulls schema and definitions from the knowledge base automatically.

Beyond answering questions on demand, Lumi is built to be proactive. Teams configure it to hunt for opportunities around objectives such as growing revenue, cutting cost, or freeing working capital, and its agents then monitor the data continuously and surface anomalies and opportunities without being asked, sending recommendations to stakeholders. The company reports customers finding seven figure opportunities within the first month, and names mid to large consumer goods and Fortune 10 retail clients including HelloFresh, GoBolt, and Chalhoub Group, plus a partnership with Deloitte. Lumi is sold as an enterprise product through a guided pilot rather than public self serve pricing, so costs are established through a sales conversation.

Vendor details

Canonical URL

https://lumi-ai.com

Category

Data analyst agent

Company status

independent

Use cases & customers

In practice

A retail pricing team asks Lumi to find margin opportunities, and its agents analyze a full year of sales and elasticity inside the company network, returning ranked SKUs without any data leaving the environment.

A procurement lead connects Lumi to the enterprise resource planning system and asks which items were bought from multiple suppliers in one month, then uses the finding to negotiate better vendor terms.

An operations team configures Lumi to hunt for working capital opportunities, and its agents monitor operational data continuously, surfacing anomalies and sending decision ready recommendations to stakeholders without waiting to be asked.

Sources & related URLs

Research notes

Score 8.5 (6F/5P/3N). Enterprise chat based MULTI AGENT data analyst for supply/distribution/retail/CPG ops (explicitly not text to SQL: multiple agents + structured KB). $3.7M seed (AgFunder, Forum, ADQ, Qora71). Clients HelloFresh, GoBolt, Chalhoub, Kroger, Growmark, F10 retail; Deloitte partnership. Fulls: Int (SAP/Oracle/MS Dynamics ERP + warehouses/DBs, real time, governed creds; executes SQL/Python), Orch (multi agent, multi step, proactive agentic workflows that monitor + act continuously, rebuilds pipeline), Know (KB: field defs, table relationships, business rules, KPIs, aliases; interprets business language), Sec (SOC 2 + GDPR + RBAC + encrypted governed connections + in network processing), Dep (runs INSIDE customer env, data never leaves, on prem AND cloud, data residency; major differentiator), Trig (proactive continuous monitoring, anomaly/opportunity alerts without human intervention, automated refresh schedules, recs to stakeholders). Partials: HITL (read only access, query cost/duration limits, RBAC, transparency; humans act on recs; no explicit approval gate), Obs (transparency into assumptions + step by step explanations + reproducible pipeline; formal audit log not confirmed), Mem (persistent KB + configured proactive objectives; conversational memory not detailed), Pack (vertical use case playbooks: pricing, procurement, supply; configured not marketplace), Eval (guided SQL/Python draft + troubleshoot + transparency; no formal eval harness). N: Model (proprietary analytics engine, no model choice/BYO documented), Ext (data gateway connectors inbound; no public API/SDK/MCP documented), Comp (no browser/computer use). Pricing sales_led / contact_only: enterprise, ~2 week pilot, white glove onboarding, no public price. Ties definite at 8.5, near lane top (buster 9.0).

Capability coverage

8.5 / 14 capabilities · 61%

Integrations & Tool CallingLumi connects to enterprise resource planning systems such as SAP, Oracle, and Microsoft Dynamics plus warehouses and databases through governed read only connectors, and generates and runs the SQL and Python to answer questions, so full. Full
Workflow OrchestrationLumi uses multiple coordinated agents to run complex multi step analyses and proactive agentic workflows that monitor data and act continuously, rebuilding the full pipeline each run, so full. Full
Knowledge Grounding & RAGLumi's knowledge base stores field definitions, table relationships, business rules, key metrics, and aliases so answers reflect the company's true business meaning, so full. Full
Human Oversight & GuardrailsLumi enforces read only access, query cost and duration limits, and role based access and sends recommendations to human stakeholders, but no explicit action approval gate is documented, so partial. Partial
Security, Identity & GovernanceLumi is SOC 2 compliant and GDPR aligned with role based access control, encrypted governed connections, and processing inside the customer network, so full. Full
Observability & AuditabilityLumi is transparent about its work, exposing every assumption and a step by step explanation of how each number was produced, but a formal audit log is not separately confirmed, so partial. Partial
Memory & State PersistenceLumi persists a configurable knowledge base and standing objectives for its proactive agents, but a first class conversational memory is not detailed, so partial. Partial
Deployment & Data ResidencyLumi runs inside the customer's own environment so data never leaves the network, and it supports both cloud and on premises deployment with data residency, so full. Full
Prebuilt Agents, Templates & PacksLumi provides configurable use case playbooks across pricing, procurement, and supply operations, but these are configured per customer rather than offered as a marketplace, so partial. Partial
Triggers & Channel CoverageLumi's proactive agents continuously monitor operational data, surface anomalies and opportunities without being asked, refresh on schedule, and send recommendations to stakeholders, so full. Full
Model Flexibility & RoutingLumi runs on a proprietary analytics engine and does not document a user facing choice of model or provider, so not documented. Unable to verify
APIs, SDKs & MCP ExtensibilityLumi connects through data gateway connectors but does not document a public application programming interface, software development kit, or Model Context Protocol surface, so not documented. Unable to verify
Testing, Debugging & OptimizationLumi offers a guided environment to draft and troubleshoot SQL and Python and full transparency into its logic, but a formal evaluation harness is not documented, so partial. Partial
Browser & Computer UseLumi analyzes structured operational data and has no browser or computer use capability, so not documented. Unable to verify

Recent platform changes

No recent material changes tracked yet.

Pricing

Custom enterprise (pilot; contact sales)

Contact onlyMedium variable cost
Verified 2026-07-01
Data confidence: medium

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.

Agentic AI Index

A directory and comparison resource for AI agent platforms, autonomous workflow tools, and enterprise agentic automation products.

© 2026 Agentic AI Index

3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746

Researched from public vendor sources. See Methodology.