LuMay AI

The exact-domain company presents LuMay as a security-first enterprise AI Agent Factory. It calls Voice, CRM, Workflow Automation, Insight, Forecasting, Legal and custom agents production products, describes action-taking across business systems, and says SOC 2 Type II and ISO/IEC 27001:2022 audits are underway rather than complete. Generic 2-4-week deployment and up-to-85% cost-reduction cards have no named cohort or formula.

The catalog identifies LegalPro+ and Voxentis as production ready and lists functional Voice, CRM, Workflow Automation, Forecasting, Insight, Translation, OCG Compliance and QMS Compliance agents, while also marking several vertical products coming soon. The status presentation conflicts with the homepage and individual pages that call a broader horizontal suite production-ready.

The platform has interface, intelligence, orchestration, integration, governance and analytics layers. It coordinates agents, tools, workflows, rules, approvals, memory, state and escalation; can trigger tasks, update systems, create tickets and follow-ups; supports voice, chat, web, Teams, email, SMS, APIs and webhooks; and offers SaaS, customer-cloud, private-cloud, hybrid, on-premises and restricted deployment plus implementation, training and managed optimization.

Voxentis handles inbound and outbound calls within approved flows and tools. It retrieves order, ticket and case state; creates or updates tickets, leads, tasks and opportunities; sends confirmations; logs outcomes; performs logistics-status and collections conversations; and transfers to a human with intent, verification, transcript and workflow context. Payment reminders and promise-to-pay capture do not prove money movement.

The ecommerce product page says Voxentis uses live OMS, order and tracking data; handles return-eligibility conversations; generates return labels through APIs; initiates refund workflows; and writes validated order, refund or ticket updates through connectors. It names Shopify, WooCommerce and Magento, although the current integration catalog and docs list only Shopify as a commerce connector.

The feature reference calls the voice capabilities production-ready and documents multi-turn within-call memory, caller recognition, multilingual operation, booking, analytics, configurable persona and restrictions, human escalation and read/write connections to CRM, calendars, ticketing and custom REST APIs. Ecommerce examples include returns/refunds and delivery support, but the page does not independently prove fund settlement.

The outbound product dials contact lists, qualifies and scores leads, routes qualified prospects, schedules callbacks, updates CRM records and runs call/SMS/email sequences. It also describes audit-ready controls and expected security-audit timing, not a public certificate.

Inbound voice reads live calendars, inventory, pricing and customer records, then can create records, send confirmations, schedule follow-ups and update systems through configured tools. It describes identity verification, restricted topics, escalation, tenant isolation, RBAC, SSO and traces; inventory and price access is read evidence rather than proof of merchant inventory or price writes.

The current catalog names Salesforce, Zoho, Freshdesk, Zendesk, ServiceNow, Twilio, Shopify, Stripe, Google Calendar, Google Sheets, Google Drive, SQL, SharePoint/OneDrive and email providers, plus REST, webhook and custom connectors. Shopify is the only currently cataloged ecommerce platform; Stripe presence does not by itself prove autonomous payment charging.

The documentation describes CRM, ticketing, Shopify, calendars, databases, files and messaging connections and custom REST or webhook integrations. It is a high-level connector guide rather than an endpoint-level public API reference.

LuMay documents a native Shopify connector for customer and order context, order lookup, tracking and return-intake or commerce workflows. No official Shopify App Store listing was found, so this remains vendor-documented connector authority rather than marketplace verification.

Connector architecture exposes create and update tools for CRM and ticketing, handles OAuth and API-key credentials, supports custom REST/webhooks, describes MCP tool discovery and records per-call OpenTelemetry traces. Public material does not expose a full developer endpoint reference or downloadable MCP server catalog.

The orchestration engine supports versioned visual or natural-language-authored flows, tool nodes, human approval gates, retries, compensation, schedules, webhooks, memory/state and multi-agent delegation. It is autonomous within configured graphs and permissions, not evidence of unrestricted open-ended authority.

Analytics cover containment, escalation, connector success, latency, cost per outcome, sentiment, work-item counts, quality, exceptions, distributed traces, logs and ROI/adoption reporting. The page explicitly says CSAT and NPS are not built in today, preserving an evaluation boundary despite broader marketing claims.

The security page describes tenant isolation, RBAC, SSO, PII masking, encryption, audit trails, guardrails and human-in-the-loop approval gates. It describes alignment and readiness; it is not a regulator or auditor certificate.

The logistics workflow looks up shipments, captures ETA or delay state, notifies dispatchers or customers, writes shipment notes and escalates failed-delivery exceptions. It does not prove courier selection, physical fulfillment, label purchasing or carrier execution.

The article explicitly claims Tamil, Hindi and Telugu dialect and code-switching support and describes Indian ecommerce shipment notifications, real-time tracking, return or exchange initiation, post-delivery surveys, batch calls and CRM actions. Its certification language conflicts with current pages saying audits are underway and is not used as certification evidence.

CRM Agents score leads, pipeline, account, quote and expansion signals; recommend next actions, cross-sell and quote-readiness changes; forecast sales; and prepare summaries, tasks and follow-up drafts. Field updates, task creation, summary attachment and action logging occur when approved, so commercial writeback is human-gated.

The Workflow Automation Agent runs configured multi-step and multi-agent work across ServiceNow, SAP, document systems, ERP and data warehouses using LangGraph, AutoGen and CrewAI. It executes unflagged steps, pauses at policy approval gates, alerts exceptions with context and preserves audit history. The current page starts at $1,499 per month; generic 11,000-hour, 92% error and 88% cycle-reduction cards have no named customer lineage.

The Forecasting Agent produces demand, workload and risk forecasts with conservative, central and stretch scenarios, confidence bands, driver decomposition and up to five years of backtesting. It refreshes automatically from connected data and is read-only by default; no downstream purchase-order or inventory write is documented. Published generic accuracy, variance, stockout and staffing cards are anonymous.

Insight Agents provide permission-aware, cited answers over SharePoint, Confluence, Jira and document repositories, with session context, query logging and knowledge refresh. This is retrieval and guidance, not autonomous system mutation.

The Anomaly Detection Agent monitors operational data and issues explanatory alerts with context. Current evidence supports detection and recommendation, not automated remediation or financial-risk decisions.

The industry index exposes ecommerce orders, returns and inquiries as a current path while marking multiple retail, logistics and other vertical packages coming soon. It is used to keep custom solution and roadmap claims separate from current shippable products.

The solution page describes demand planning, stockout and disruption signals and India DPDP alignment for custom supply-chain deployments. It supports forecasting and advisory fit but does not prove a generally available inventory-execution system or a named merchant customer.

The main commercial page starts Voice at $179/month, Translation at $499, Insight, CRM, Forecasting and Anomaly at $999, Workflow at $1,499, Legal at $7,500 and OCG/QMS at $10,000. Custom pilots start at $25,000; customer-cloud starts $36,000/year, on-prem $90,000/year plus $75,000 setup, air-gapped and Enterprise Platform $150,000/year. It publishes Discovery Sprint, Production Pilot, Engineering Pod, Managed Optimization and Academy prices, monthly/annual billing and volume/integration/deployment/security quote drivers.

The dedicated page contradicts the main pricing authority: it displays Basic $1/month for 200 minutes, Starter $2.50 for 500, Growth $7.50 for 1,500, Scale $25 for 5,000 and Enterprise $50 for 10,000, while separately saying $0.05 per minute, $0.005 overage and $0.05-$0.10 per-minute connectors. It includes a 14-day trial and mentions REST, webhooks, GraphQL and MCP servers. The inconsistent live values prevent a single definitive voice entry price.

LuMay describes a services boundary spanning strategy and advisory, implementation, AI engineering, managed optimization, support and training. These human-led activities configure and operate the platform but are not a managed human call-center workforce.

The story page says LuMay was founded in 2025 or Q1 2025, has 15+ team members and reports 50+ enterprise clients, 10 agents, 75% average cost reduction, 90% productivity improvement and 2-4-week deployment. These vendor aggregates lack cohorts/formulas and the founding and team claims conflict with official LinkedIn.

The page identifies Palani Ramasamy as CEO and Founder and names Harun Diri, Mike Millard, Mary Grygleski, Prashant Bhuptani, Surekha Chikoti and Reshma Sriram in engineering, strategy, market, delivery, services, product, partnership and enablement roles.

The contact page names LuMay INC and gives 8 The Green #20160, Dover, Delaware 19901, United States, sales@lumay.ai and a US telephone number. This is company-claimed identity and address, not a regulator-confirmed incorporation record. The page says inquiries receive a response within one business day.

The exact-domain official profile says LuMay is privately held, headquartered in Dover, founded in 2024 and in the 51-200 company-size band. Current posts recruit in Chennai, Trichy and remote India and announce interns, proving an India hiring footprint but not an India-headquartered or India-first company. The founding year and team band conflict with the company story page.

Terms last updated 10 February 2026 describe LuMay AI services for enterprise and business use, customer ownership of customer data, a processing license, third-party service dependencies and as-is availability. They say CRM data is not retained beyond the active session and select California governing law, but do not provide a registration number or cap table.

The policy covers enterprise-client prompts, files, logs, telemetry and billing data, cross-border transfers and enterprise DPAs, and says CRM data is transient rather than retained. Its displayed update chronology is inconsistent and is not used to date the policy.

The careers page describes a remote-first company and current US-based role language. Together with official social posts recruiting in India, it supports a distributed operating footprint rather than an India headquarters claim.

The enterprise guide targets CIO, CTO, general counsel and chief data officer audiences and frames governed agent deployment for enterprise buyers. Exact economic buyer, procurement cycle and conversion rate remain undisclosed.

Microsoft Marketplace lists a LuMay INC Smart Translation SaaS offer, providing third-party official-marketplace confirmation of the exact LuMay entity and one product. It does not verify Voxentis, Shopify distribution, customer outcomes or the full product catalog.

LuMay's current case index exposes four detail stories: EDI Translation Agents, a financial-services Voice Agents deployment associated with Reema, a QMS story whose headline names goxpedite, and the internal MyLu build. The index reports 12 hours to 2 minutes and 85% cost reduction for EDI; 18 minutes to under 4 minutes and 55% higher CSAT for the voice story; 60% audit-prep reduction and 3x audit throughput for QMS; and a five-day internal MyLu build. Several index values conflict with the linked detail pages.

The summary repeats EDI, financial-services voice and QMS outcomes and claims every metric is named and methodology-cited. In practice the voice customer company remains anonymous, and the QMS identity and values conflict across LuMay surfaces. It adds EDI expansion from 3 to 14+ languages; voice results of +18 CSAT points, 62% lower annual operating cost and 1.2 million calls per month; QMS audit findings down 87%; first-agent delivery in 2–4 weeks; and ROI-positive results by week 12 for most engagements, with one EU on-premises outlier taking longer.

LuMay attributes to EDI and Samy Rajendran an 85% reduction in annual translation cost, processing of a 140-page document from 12 hours to 2 minutes or under 2 minutes, 90% automation from a fully manual baseline, 99.99% accuracy, 500,000+ pages per year without manual intervention, and 60%+ revenue-cycle improvement. It calls 12 hours to 2 minutes a 360x throughput increase. No customer-owned source confirms these outcome values, and the revenue and accuracy formulas are not disclosed.

An anonymous North American financial-services team is described as replacing after-hours voicemail with Voxentis. LuMay reports +18 CSAT points by month 2, 62% lower annual call-center operating cost, 1.2 million calls per month, a stated peak day of 50,000 concurrent calls, 43% lower average handle time, and expansion from English to three languages in week 4. Reema Rochelle is quoted and marked verified, but her employer is not named. The case index separately says 18 minutes to under 4 minutes and 55% higher CSAT, which cannot be reconciled to the detail page's 43% and +18-point values.

The headline names goxpedite, while the body describes a leading medical-device compliance firm and attributes a quote to Markus Hoffmann at a German medical-device manufacturer. The same page reports 2–3 weeks to under 4 hours and labels this 95% faster; 60% team productivity gain; document-retrieval time falling from 70% to 20% of engineer time; 90% automation from 0%; 4 to 12+ audits per quarter; and 200+ QMS documents checked in under 4 hours. The case index instead says audit prep fell 60%, and the success hub calls the customer MedDevice and adds an 87% reduction in audit findings.

LuMay documents its own five-day build of MyLu from sketch to production launch. The page reports 40% faster execution, 70% less manual search and 24/7 availability without disclosing baselines, sample sizes, measurement windows or formulas. This is an internal product-build narrative, not external customer confirmation.

A legal-industry page labels three values as production results from cited legal customers: 90% lower contract-review time, 100% OCG compliance coverage and 40% fewer invoice disputes. An anonymous Fortune 500 GC-office counsel lead is quoted saying a redacted $XXX,000 of non-compliant billing was found in the first quarter. No customer name, source citation, baseline amounts, cohort or calculation method is public.

A healthcare page reports 47% higher intake completion, 78% shorter prior-authorization cycle time and 22% higher follow-up adherence as production results from cited health-system customers. An anonymous chief nursing informatics officer at a 12-hospital regional system says patient NPS rose 18 points after Voice Agents handled routine calls. The page does not identify the health system or link a customer-owned confirmation, and it does not establish that every displayed metric belongs to that one testimonial customer.

A finance page presents 71% fewer false positives, +18 CSAT points and 31% lower forecast variance within two quarters as production results from cited financial customers. The +18 CSAT value explicitly points to the separate Voice Agents case. An anonymous marketplace trust-and-safety leader for a platform with 80M+ users is quoted saying alert volume fell from thousands per day to under 20. The remaining values are not tied to a named customer or customer-owned source.

A supply-chain page reports 31% lower forecast variance within two quarters, 64% fewer stock-out incidents and disruption warning compressed from days to hours. It carries an anonymous quote attributed to the head of FP&A at a multi-billion-dollar retailer. The source does not name the retailer, link a customer confirmation or disclose baselines and formulas, and it does not explicitly prove that all three values belong to the testimonial customer.

The hub names Sutherland Global, Gategroup, Cyber Assuranze, two anonymous global law firms, LuMay India Private Ltd, Ecube Consulting and HD Soft Consulting. It describes deployments or reseller rollouts, reports more than 25% lower incident-response costs for Cyber Assuranze, and says Aon pilots are underway through an APAC rollout. The destination links are organization homepages rather than customer confirmations of LuMay use.

LuMay says it delivered corporate AI, secure-development and governance training for EDI. This corroborates a current LuMay–EDI relationship but does not customer-confirm the Translation Agents outcomes in LuMay's case study.

An Exchange Data International document identifies Samy Rajendran as Director, International Business Strategy. It helps resolve the EDI and spokesperson identity used in LuMay's case, but it predates LuMay and does not confirm the LuMay deployment or any outcome metric.

LuMay describes an anonymous global manufacturing enterprise moving from concept to full production in under four weeks with compliance and governance active on day one. The quoted CTO is unnamed and no workflow scope, start date, customer source or measurement method is disclosed.

Microsoft Marketplace lists LuMay INC's Smart Translation offer and exposes a Ratings + reviews tab, but no public rating aggregate or review text was rendered at access. The listing's generic 95% faster, 275% cost reduction and 99.99% accuracy statements are product benefit claims, not attributable customer cases, and are excluded from case normalization.

Microsoft Marketplace lists LuMay INC's Dynamics 365 Agentic AI offer and exposes a Ratings + reviews tab, but no public rating aggregate or review text was rendered at access. Its 70% effort savings, 99.7% accuracy, 95% adoption and 80% faster onboarding claims are unattributed product benefits and are not normalized as customer cases.

The sitemap inventories the current resources case index, resources success summary, four detail pages, a legacy /about/successtories alias, a SmartCall alias that redirects to the Voice Agents case, and design-preview routes. Design-preview pages are internal template demonstrations rather than distinct customer evidence; the SmartCall and about aliases duplicate current canonical pages.

LuMay says it has served more than 50 enterprise clients and reports 75% average cost reduction, 90% productivity gain and 2–4 week average go-live. A timeline attributes 75% cost reduction and 90% productivity gains to its first 2025 deployments across logistics, healthcare and financial services. The page does not identify the relevant customers, denominators, weighting method or measured workflows.