FAQs
Quick answers — then the full glossary below if you need definitions.
MaxxBot is an AI procurement assistant grounded in structured data — not generic AI responses. Every answer shows its source and a confidence level. It is currently in Beta.
MaxxBot queries structured procurement tables, not the open internet. Every answer is sourced from a specific table or methodology section. It does not hallucinate spend numbers.
MaxxBot supports vendor intelligence, rate benchmarking, managed-services pricing builds, contract review, RFP support, taxonomy navigation, and Premium command-center analytics on your classified data. Each response is labeled by source type and confidence.
414 service types across 94 categories and 11 towers — the structured knowledge base MaxxBot draws from. Browse it on the Taxonomy page.
Depth and specificity. Most procurement tools classify spend into 5–15 generic categories. The Maxx taxonomy has 414 service types across 94 categories and 11 towers — so "IT spend" becomes Cloud Hosting & IaaS, Data Center Operations, Cybersecurity Consulting, and other distinct, queryable service types.
Benchmarks are always ranges, never single points. Each response includes a confidence level (High, Medium, or Low) and a source citation. MaxxBot tells you when a figure is estimated and recommends validation before acting on low-confidence answers.
No. Standard tier works with the public taxonomy and benchmark tables. Premium — MaxxBot on your classified spend — requires a completed Assessment.
Classification is part of the Maxx Assessment and Command Center workflow (Premium). Maxx maps invoice rows using vendor identity, purchase context, and methodology rules, then validates assignments with your team before operational use.
MaxxBot says so. It will not fabricate an answer. It will report low confidence or that the question falls outside its current data, and suggest where to look next.
Yes on Premium: MaxxBot connects to your classified spend cube (vendor scorecards, category analytics, PO compliance, trends). Standard tier uses the shared taxonomy and benchmarks only.
For Premium, raw invoice rows do not enter the bot — only pre-aggregated summaries. Workspaces are isolated per client. Enterprise LLM API terms prohibit using your data for model training. Data is encrypted in transit (TLS 1.2+) and at rest (AES-256). Authentication uses Clerk (SOC 2 Type II). Hosting is on Vercel (SOC 2 Type II, ISO 27001). ScaleX SOC 2 is in progress.
Standard tier: taxonomy and benchmark updates are periodic. Premium: monthly refresh aligned with your Command Center cycle.
Request early access. After April 15, sign up for a free one-week trial. To upgrade to Premium, contact us about a Maxx Assessment.
Yes. We actively accept service type requests, description improvements, and coverage gap reports from the procurement community. Accepted contributions earn an extra week on your free trial. Submit feedback on the Taxonomy Explorer page.
Glossary
Key terms used across taxonomy and MaxxBot responses.
- Tower
- The highest level of the Maxx taxonomy — the functional domain that owns the spend. Active towers include: Supply Chain, Customer Operations, Marketing & Growth Services, Information Technology, Risk & Compliance, Facilities & Real Estate, Engineering & Product Development, Finance & Accounting, Content Operations, AI & Data Services, and Human Resources.
- Category
- The second level of the taxonomy, grouping related spend areas within a tower. There are 94 categories. Example: within IT, "Infrastructure & Cloud Operations" is a category covering cloud hosting, data centers, and network services.
- Service Type
- The third level — the specific service being purchased. There are 414 service types. Example: within IT → Infrastructure & Cloud, "Cloud Hosting & IaaS" is a service type.
- Workflow
- The fourth and most specific taxonomy level — the engagement or project context. Workflows are client-specific and defined during the Assessment. Example: Cloud Hosting & IaaS → "AWS production migration."
- Spend Nature
- A classification of what an invoice row represents economically. Six values: Third-Party Vendor (addressable procurement spend), Direct Payment (client-to-employee or statutory payments), Employee Withholding (payroll deductions), Internal Cross-Charge (intercompany allocations), Bonus Payment (variable compensation), Regulatory Compliance (government fees and filings).
- Addressable Spend
- The portion of total spend where procurement can influence cost, terms, or vendor selection. Excludes pass-throughs, withholdings, internal charges, and regulatory payments.
- PO Compliance
- The percentage of spend covered by a purchase order. A measure of procurement governance — higher PO compliance means more spend is pre-approved and tracked against contracts.
- Vendor Consolidation
- The practice of reducing the number of vendors serving the same service type or category, typically to strengthen negotiation leverage and simplify management. The parent company heatmap in the Assessment reveals consolidation opportunities.
- Parent Company
- The corporate parent that owns a vendor entity. Multiple vendor names in AP data often roll up to the same parent — a consolidation opportunity invisible without parent mapping.
- Advisor Flag
- A classification marker indicating whether a vendor provides advisory services (consulting, audit, legal counsel) as opposed to operational or transactional services. Used for spend analysis and governance segmentation.
- Cost of Revenue
- Spend that directly supports the client's revenue-generating operations (product delivery, service fulfillment) as opposed to general and administrative functions. Identified through functional area analysis and cost center signals.
- Shadow IT
- Technology purchases made by business units outside the CIO's organizational purview. Identified when IT tower spend appears under non-IT functional areas or cost centers. A common finding in enterprise assessments.
- Taxonomy
- The complete hierarchical classification system used to organize spend data. The Maxx taxonomy has 4 levels: Tower → Category → Service Type → Workflow.
- Classification
- The process of assigning each invoice row to its correct tower, category, service type, and spend nature within the taxonomy.
- Benchmark Range
- A quantitative reference range for a service type or category, presented as a low-to-high band with a source citation and confidence level. Maxx never presents benchmarks as single-point figures.
- Confidence Level
- An indicator of how certain a response is: High (direct table match from your data), Medium (derived from related data), or Low (estimated from methodology). Shown on every MaxxBot response that includes a quantitative claim.
- Assessment
- The initial engagement phase where AP/invoice data is classified against the Maxx taxonomy, producing a spend cube, interactive report, and command center workbook.
- Command Center
- The monthly operating deliverable — an Excel workbook tracking spend, targets, vendor metrics, and governance indicators. Updated each cycle with the latest data.
- MaxxBot
- The conversational interface to classified spend data and the shared taxonomy. Queries structured tables and provides sourced, confidence-rated responses. Currently in Beta.
- Spend Cube
- The structured output of classification — every invoice row mapped to tower, category, service type, vendor, spend nature, and organizational attribution. The foundation for all Maxx analysis and reporting.