General-business eval scoreboard

SLAtech AI Business: 84/100

Reproducible 200-question general-business eval harness. SLAtech AI Business scores 84/100 against a 71/100 generic baseline used to compare all per-vertical specialised models. Pairs with umbrella eval scoreboard, general-business glossary and general-business FAQ.

Score breakdown by category

Category	SLAtech AI Business	Generic competitor	Gap
Lead-capture quality Structured name + company + role + use-case intake. Generic chatbots collect contact form fields only - no qualification depth.	88	68	+20
General FAQ coverage Hours / pricing / policy queries hit cached answer cards. Generic chatbots match here when given clean FAQ source.	85	71	+14
Routing to specialist vertical Self-classification: identifies industry-vertical from query and routes to the relevant Med/Edu/Legal hub. Generic chatbots have no vertical-routing intelligence.	86	52	+34
Multilingual support (HE / RU) Native English, Hebrew (full RTL) and Russian with per-message language detection. Generic chatbots lean on broad auto-translate without RTL polish or native terminology.	82	70	+12
Confidentiality / PII posture PII redaction at ingest with tenant isolation. Vertical models (Med, Legal) ship deeper redaction + BAA/UPL workflows.	79	60	+19

84/100

Own June-2026 eval score; vertical-routing native, RTL Hebrew polish. Generic baseline reference = 71/100.

73/100

Mid-market polish but English-first, no vertical-routing, no eval published

70/100

SMB-priced but no Hebrew RTL polish, conversation cap on lower tiers

68/100

Popular SMB bot-builder but English-first, no vertical-routing, no Hebrew RTL polish

67/100

Strong demo-booking but English-first, US-locale defaults

The per-vertical eval score is one input. Three more self-serve tools complete the picture without a sales call:

Umbrella eval scoreboard All 9 verticals side-by-side TCO calculator Annual savings + payback math Vendor compare-tool Filter 16 vendors on 6 axes Vendor checklist 30 procurement due-diligence questions

Eval methodology is open-source. 200 sealed general-business questions with LLM-as-Judge scoring on factuality, hallucination and confidence axes.