Medius vs Vic.ai

Requirement evaluated: Segregation of duties enforcement: person who enters cannot approve, person who approves cannot process payment

For a 3-person AP team at a multi-location services company, Medius enforces all three legs of segregation of duties through distinct, system-level mechanisms rather than policy-only controls. First, the `AllowInspectAndAttest` system parameter controls whether the user who reviewed and coded an invoice can also approve it: <cite index="33-15">setting this to 'No' means that a person who has reviewed a coding row can never approve the same row.</cite> Second, Medius includes a configurable Four Eyes Principle (4EP) …

Limitations: All three controls (AllowInspectAndAttest, Four Eyes Principle, Pay Approver Role) are configurable settings that must be deliberately activated during implementation; they are not all enabled out-of-the-box, so the buyer's implementation team must confirm each is turned on and that no threshold-based bypass exists (fo …

Requirement evaluated: Segregation of duties enforcement: person who enters cannot approve, person who approves cannot process payment

For a 3-person AP team moving from manual email-chain approvals to a system-enforced control environment, Vic.ai's role architecture creates the three-stage separation your policy requires. The platform distinguishes an Accountant role (invoice entry, AI-prediction review, and coding on the Invoices tab) from a separately designated Approver role, and from payment execution permissions managed through VicPay. …

Limitations: One configuration item to lock down: the help center notes that 'Accountant users can modify the approval flow as needed before starting them,' which means an Accountant with broad access could redirect a flow before initiating it — the admin should restrict approval flow editing rights and set Autonomous Approval Flow …

Requirement evaluated: Cash flow forecasting based on approved and pending payables with due date distribution

For a $120M multi-location services company currently flying blind on payables timing, Medius delivers a dedicated Cashflow Dashboard within its Medius Analytics module, listed as a named pre-built dashboard in the Medius help center documentation alongside the Overview, Capture Insights, and Straight Through Processing dashboards. The mechanism works because Medius sits in the pre-processing layer: every invoice, whether still in an approval queue or fully approved and awaiting payment, lives in the Medius pipeline with its due date, payment terms, and workflow status tracked in real time. …

Limitations: The Medius Analytics layer, which houses the Cashflow Dashboard, refreshes on a Medius-managed schedule rather than continuously, so the view is a periodically updated snapshot rather than a second-by-second live feed; the operational gadget dashboards do update in real time but are invoice-status views rather than for …

Requirement evaluated: Cash flow forecasting based on approved and pending payables with due date distribution

For a 3-person AP team currently flying blind on cash timing, Vic.ai's VicAnalytics module and VicAgents layer offer the closest match to this requirement. VicAnalytics explicitly lists 'trustworthy accruals and forecasting' as a feature, framed as financial planning built on 'real, up-to-date AP data for better cash flow management' — and because Vic.ai captures invoice due dates at the header level during processing (before posting to Sage Intacct), that data pool includes invoices still moving through the approval queue, not just posted transactions. …

Limitations: For this buyer's specific use case — seeing which approved invoices are due in the next 30 days versus which pending invoices are still in approval and projected to come due — Vic.ai's documented analytics appear to stop at high-level cash flow trend and accruals visibility rather than a purpose-built due-date distribu …

Requirement evaluated: The AI coding model must learn from this buyer's specific transaction history to improve dimension coding accuracy over time, using the 12,000 monthly invoices as the training corpus. The vendor must explain the actual mechanism (per-customer model, fine-tuning on approval history, rules derived from prior accepted coding, or equivalent) and must not describe a generic pretrained model as if it were customer-specific learning. The buyer's question, 'how does the per-customer model learn from our history,' must be answerable with a concrete mechanism and a measurable lift curve, not a marketing claim.

For a buyer processing 12,000 invoices a month across dozens of NetSuite coding fields, Medius's learning mechanism is delivered through SmartFlow, a CNN-based proprietary model that auto-codes GL account, tax fields, approver values, and coding dimensions for non-PO invoices. The mechanism is explicitly company-specific: according to Medius's invoice automation product page, SmartFlow is 'trained on your historical actions and enriched by 2.4 billion+ invoice field data points across Medius's global customer base,' and a Medius Chief Architect confirmed in a published interview that 'our machine learning technology uses pattern recognition to capture invoices, code them correctly, and route …

Limitations: Medius does not publish a measurable lift curve showing accuracy improvement as a function of invoice volume past the cold-start '95% after two invoices' benchmark, so the buyer cannot verify the concrete progression their 12,000 monthly invoices would drive. …

Requirement evaluated: The AI coding model must learn from this buyer's specific transaction history to improve dimension coding accuracy over time, using the 12,000 monthly invoices as the training corpus. The vendor must explain the actual mechanism (per-customer model, fine-tuning on approval history, rules derived from prior accepted coding, or equivalent) and must not describe a generic pretrained model as if it were customer-specific learning. The buyer's question, 'how does the per-customer model learn from our history,' must be answerable with a concrete mechanism and a measurable lift curve, not a marketing claim.

For a buyer processing 12,000 invoices a month on NetSuite with dozens of coding fields, Vic.ai's approach starts before go-live: at onboarding, the platform ingests the buyer's historical approved invoices as a dedicated training corpus, and the vendor's own API documentation exposes specific endpoints to sync those historical invoices into 'your AI model' for pre-training (Vic.ai API docs). Once live, every AP staff confirmation or correction becomes a labeled training signal: the AI makes predictions at both the header level (invoice number, date, amount, currency) …

Limitations: The buyer's requirement for a provable per-customer model that is isolated from all other customers' data is not definitively answered by Vic.ai's published documentation: one technical analysis characterizes the architecture as a global multi-tenant model that improves across all clients simultaneously from anonymized …

Requirement evaluated: International wire payments to 8 overseas vendors with multi-currency support

Your AP team currently generates international wire instructions manually outside Sage Intacct for 8 overseas vendors. With Medius Payments (the vendor's payment execution module, priced separately from core AP automation), once an invoice completes the Medius approval workflow, a wire payment is formatted within Medius using the vendor's stored SWIFT/BIC code or IBAN, then transmitted via secure FTP directly to your financial institution's wire desk for settlement — no manual Sage Intacct export or separate bank login required. …

Limitations: For US-based companies, the international wire execution model routes formatted wire instructions through your existing financial institution's wire desk via secure FTP rather than through Medius's own direct payment rails, so settlement speed and any per-wire fees remain subject to your bank's international wire polic …

Requirement evaluated: International wire payments to 8 overseas vendors with multi-currency support

For a $120M services company needing to pay 8 overseas vendors in multiple currencies, Vic.ai's VicPay module does list international wires as a supported payment method alongside check, ACH, and virtual card, all within the same platform workflow after invoice approval. Once an invoice completes the approval stage, a Payment Initiator assembles a payment batch, selects the pay-from bank account, and submits for a Payment Approver to release; the payment data then syncs back to the connected Sage Intacct entities via a reconciliation API. …

Limitations: For this buyer's 8 overseas vendors, the one-currency-per-batch constraint means international wire payments cannot be consolidated into a single payment run alongside domestic disbursements, adding manual batching steps and increasing processing time. …

Requirement evaluated: The system must support full NetSuite custom segment coding, not only the standard NetSuite dimensions (GL account, location, department, class, project). The buyer explicitly calls out 'several custom dimensions' as part of their standard coding workflow. A vendor whose data model is limited to NetSuite's out-of-the-box fields cannot serve this buyer; the integration layer must read the buyer's NetSuite custom segment configuration and expose those segments as codeable targets in the AP automation UI and AI coding engine.

For a buyer running NetSuite with dozens of coding fields including several custom dimensions, Medius addresses the requirement through a direct schema import from NetSuite at onboarding. The May 2025 Medius AP Automation for Oracle NetSuite product definition states explicitly: 'The structure of coding dimensions, including both standard and custom dimensions, is determined during the data gathering phase of the customer onboarding process. …

Limitations: The product definition document describes the custom dimension import as part of the customer onboarding process, which implies configuration effort at implementation rather than a self-service, real-time sync; if the buyer adds new custom segments to NetSuite post-go-live, it is unclear from available documentation wh …

Requirement evaluated: The system must support full NetSuite custom segment coding, not only the standard NetSuite dimensions (GL account, location, department, class, project). The buyer explicitly calls out 'several custom dimensions' as part of their standard coding workflow. A vendor whose data model is limited to NetSuite's out-of-the-box fields cannot serve this buyer; the integration layer must read the buyer's NetSuite custom segment configuration and expose those segments as codeable targets in the AP automation UI and AI coding engine.

For a buyer coding dozens of fields per invoice across standard and custom NetSuite dimensions, Vic.ai's Autopilot AI operates at the pre-processing journey's coding stage: it ingests invoices, then predicts both header-level data (invoice number, due date, amount, currency) and line-level data before routing to approvers. Vic.ai's NetSuite integration page describes the AI as classifying 'cost accounts, dimensions, assets' and pushing 'all the associated coding' back to NetSuite via real-time bi-directional sync. …

Limitations: The buyer's requirement is specifically for NetSuite custom segments beyond the standard field set (GL account, location, department, class, project), and no available Vic.ai documentation confirms that the integration automatically ingests and exposes arbitrary cseg_ fields as AI-codeable dimensions. …

Requirement evaluated: Vendor self-service portal: new vendor registration, W-9/W-8 submission, banking detail entry, invoice submission, payment status inquiry

For a $120M services company currently managing vendor onboarding entirely via email and manual data entry into Sage Intacct, Medius offers a dedicated, supplier-only self-service portal that covers all five sub-requirements in the buyer's ask. On the registration and onboarding side, <cite index="4-1,4-13">a dedicated self-serve portal gives suppliers the flexibility to respond to onboarding forms from all of their customers in one place; it is vendor management software that puts the onus on the supplier to onboard and maintain their details in a secure, self-service online portal.</cite> The buyer's AP team creates and issues onboarding forms in Medius Supplier Onboarding; <cite index="24 …

Limitations: Medius's onboarding questionnaires are configurable for document upload including tax forms, but no Medius documentation explicitly confirms a native W-9/W-8 collection workflow with IRS TIN matching or automated validation; buyers with high volumes of subcontractors who need structured, validated tax-form collection s …

Requirement evaluated: Centralized vendor master synchronized bidirectionally with Sage Intacct

For a multi-location services company running two Sage Intacct entities, Vic.ai's Sage Intacct Marketplace integration automatically pulls vendors, GL accounts, and dimensions from Sage Intacct into Vic.ai, making that data available for AI-driven coding and vendor predictions during invoice processing. Once an invoice is approved in Vic.ai, it is posted back to Sage Intacct's AP module with its coding intact. Within Vic.ai, the vendor information card surfaces masterdata drawn from the connected Sage Intacct account, and the AI's prediction accuracy depends on how closely vendor names in that masterdata match what appears on incoming invoices. …

Limitations: The documented sync is inbound: vendor records originate and are maintained in Sage Intacct and flow into Vic.ai. If your AP team encounters a new vendor during invoice processing and needs to add it, the current evidence indicates that vendor would need to be created in Sage Intacct first before it is available in Vic …

Requirement evaluated: Learning capability: accuracy should improve over time on our specific vendor invoice formats

For a 1,800-invoice-per-month services company starting from zero automation, Medius addresses this requirement through two complementary mechanisms in its invoice capture stage (pre-processing stage 1: legitimacy and initial data extraction). First, Medius Capture uses a proprietary multi-stage AI pipeline combining Siamese CNNs for document classification and Markov models for line-item extraction, trained on a global corpus of 2.4 billion+ invoice field data points including 393 million real-world human corrections across its customer base. Second, and directly relevant to per-vendor format improvement, SmartFlow (a proprietary CNN) …

Limitations: The precise boundary between global cross-customer model retraining and this buyer's tenant-specific model is not fully disclosed in public documentation; accuracy improvement on genuinely novel or low-volume vendor formats depends on correction volume from this buyer's own invoice corpus, and very infrequent suppliers …

Requirement evaluated: Learning capability: accuracy should improve over time on our specific vendor invoice formats

For your 3-person AP team processing 1,800 invoices per month across two Sage Intacct entities, Vic.ai's learning mechanism operates on two layers that directly address your requirement. First, the platform's AI starts with a foundation trained on over one billion invoices globally, which means it can read and extract data from your vendors' invoice formats from day one without requiring any template setup per vendor. …

Limitations: One dimension to monitor during implementation: Vic.ai's continuous learning also draws on anonymized, cross-customer data from across the platform, so some accuracy gains are shared broadly rather than being exclusively derived from your own correction history. …

Requirement evaluated: Multi-entity support within the integration; we operate 2 entities in Intacct and plan to add a third

As a $120M multi-location services company operating 2 Sage Intacct entities with a third planned, you need an AP layer that can route invoices to the right entity-scoped workflow, post to the correct entity's GL, and scale to a third entity without re-implementation. Medius's own platform supports multi-entity environments through a 'company' construct with entity-scoped accounting templates, entity-aware approval routing, and virtual-company-level analytics visible from a single login, as documented in the Medius Success Portal. …

Limitations: The Sage Intacct connector is partner-delivered via Acuity Solutions, not a Medius-native integration, and its per-entity GL mapping, Intacct dimension carriage, and multi-entity field fidelity are not publicly documented; the buyer must verify these specifics with Medius and Acuity before contracting to confirm the co …

Requirement evaluated: Integration setup assistance included in implementation; not a separate SOW or additional cost

For a multi-entity Sage Intacct customer like yours, Vic.ai does describe ERP integration as a component of its standard onboarding process. <cite index="16-1">The stated goal of Vic.ai's AI onboarding is to ensure the system understands historical data, vendor list, chart of accounts, approval flows, and that it integrates with the customer's ERP tool.</cite> <cite index="14-9,14-10">Vic.ai describes itself as designed to minimize IT involvement, with most implementations led by finance teams alongside Vic.ai onboarding specialists who provide training and configuration assistance.</cite> The ERP integrations page similarly states <cite index="21-8,21-9">'Our team configures, tests, and tra …

Limitations: No publicly available Vic.ai documentation, pricing page, or help center article explicitly states that Sage Intacct integration setup is included in the implementation cost and not billed as a separate SOW. …

Requirement evaluated: Data encryption at rest and in transit

For a $120M multi-location services company handling invoice data, vendor credentials, and payment information across two Sage Intacct entities, Medius provides encryption controls at both the storage and transmission layers. On the storage side, Medius explicitly documents AES-256 encryption across its infrastructure, which is hosted in Microsoft Azure data centers with customer data separated into unique SQL databases per customer. On the transmission side, Medius's Trust Center confirms that Transport Layer Security (TLS) …

Limitations: Medius's publicly accessible Trust Center pages confirm AES-256 and TLS but do not enumerate the specific TLS version floor (1.2 vs. 1.3) in free-text form; buyers with contractual requirements for a minimum TLS version should request the full SOC 2 Type 2 report and Qualys detail from trust.medius.com to confirm. …

Requirement evaluated: Complete audit trail: every action timestamped with user ID, viewable by invoice or by user

For a three-person AP team processing 1,800 invoices per month across two Sage Intacct entities, Vic.ai provides a dedicated Audit Log embedded in the side drawer of every invoice in the platform. From the moment an invoice is ingested, every change is recorded and attributed to a named user with a timestamp, covering the full pre-processing journey: ingestion, AI coding, human edits, approval actions, comments, rejections, and posting. The invoice-centric view is accessed directly from the invoice grid by clicking into the Status column, giving AP staff and auditors a chronological event feed for any individual invoice without leaving the document context. …

Limitations: The documented mechanism is strongest for the invoice-centric view: the help center article describes the audit log as residing in the side drawer of each invoice, with no explicit documentation of a standalone cross-invoice report filterable purely by user ID (i.e., 'show me every action taken by AP Clerk A across all …

Requirement evaluated: Two-way matching for service POs where no goods receipt applies

For a multi-location services company with subcontractor, facilities, and professional-services POs, Medius operates at pre-processing stages 2 and 5 (PO match and cost allocation) and explicitly supports 2-way matching as a configurable match type for service-based purchases. The Medius glossary directly answers the buyer's scenario: 'Can invoice matching be tailored for service-based purchases without a goods receipt? Yes. …

Limitations: The product definition documentation confirming the 2-way matching policy is drawn from MediusFlow/D365 integration specs dated 2017-2018; configuration options for the current Sage Intacct connector should be verified with Medius directly to confirm that the 2-way match policy is available and configurable at the PO-t …

Requirement evaluated: Automated three-way matching: invoice to PO to goods receipt, with configurable tolerance (2% price, 5% quantity)

For a $120M services company running 55% PO-based invoices across facilities, supplies, and subcontractors, Vic.ai's Autonomous PO Matching module handles pre-processing stages 2 through 4: PO line matching, terms verification, and receipt confirmation. The AI extracts line-item data from each invoice and compares it against the corresponding PO lines, evaluating quantities, unit prices, and descriptions; and, when a goods receipt is present, performs a full three-way match against the receipt as well. …

Limitations: The help center documents that tolerance configuration operates at the organization and company level; there is no public documentation confirming that separate tolerance rules can be set per vendor, per location, or per PO line category (e.g., 2% price tolerance for supplies vs. …