Sensing

The offer IS the AI system — a SaaS platform generating regulatory-compliant BIM models for residential feasibility in real-time, with a compounding proprietary dataset of 150M+ sqm generated across 140+ countries.

SaaS platform play with proprietary Halo Intelligence® AI layer — the offer IS the system, not consulting wrapped around it, with Gartner 4x Leader validation and quantified client outcomes.

Proprietary 750k-schedule data moat in construction project controls — the offer IS the AI system, and each engagement compounds the forecasting accuracy for all clients.

Life sciences R&D platform with a proprietary 500M-fact scientific knowledge graph and agentic orchestration — the data moat compounds with each enterprise integration, not a consulting wrapper.

The offer IS the AI system — a purpose-built finance platform processing 1.5B pages with workflow automation that encodes firm processes, scoring maximum on AI locus and system depth.

The offer IS the AI system — a legal-domain model trained on 142,000+ professional users across 1,500+ firms, creating a compounding data moat that widens with every engagement.

AI-native RCM-as-managed-service for ambulance billing — operator model with gainshare alignment, acquiring a legacy agency as distribution, revenue-per-FTE doubling as the proof point.

EHR-native AI agents that run primary-care VBC workflows end-to-end — the offer is a managed operating capability, not consulting, with a wedge sharp enough to name the buyer's contract type and current pain.

Published per-document flat pricing with real dollar amounts, a '80% AI / 20% lawyer' architecture, and same-day turnaround — the pricing model and delivery stack are both structurally novel for a law firm.

studio-product-hybrid — explicit "agency" framing combined with product-shaped delivery (the agents do everything). Could also argue productised-operator; the "AI-Native Marketing Agency" branding tips it toward hybrid.

DRAFT — Vertical SaaS-as-infrastructure play: Labrador runs the full publishing stack for 386 newsrooms across 21 countries with AI embedded as native capability, not bolted-on consulting.

Not a consulting offer — a productised AI rendering SaaS with 580k users; structurally differentiated as a software product, but misfit for this consulting rubric.

Not a consulting offer at all — a fully productised AI SaaS with published pricing, API, and 1M+ rooms of compounding training signal; the rubric barely applies.

Solo-built SaaS product (levelsio) where the offer IS the AI system — consulting rubric barely applies; structurally a product company, not a consulting firm.

SaaS-not-consulting: the shoppable real-furniture linkage turns AI staging into a commerce layer, with $50/mo flat pricing published alongside a full competitor comparison table.

PropTech-construction hybrid where the software (TESSA) algorithmically generates engineered housing schemes and the physical Kit of Parts is the proprietary supply chain — the consulting wraps a genuine product stack.

800-person Polish delivery firm with named proprietary products (futureClaims™) and a stated pay-for-performance model — credible differentiation if the gainshare mechanism is real, but no public evidence to verify.

Vertical AI product — not consulting — purpose-built for construction's unstructured data problem, with two named modules and an agentic workflow layer embedded in client operations.

Proprietary Language² architecture with cross-client performance data moat; the platform learns across all brand engagements, making each deployment structurally sharper than the last.

The offer IS the AI system — Decagon sells a fully managed omnichannel agent platform with proprietary AOP language, positioning it structurally closer to infrastructure than consulting.

Platform-native AI agent OS with outcome-based pricing and a self-improving 'agent-building agent' (Ghostwriter) — the product IS the consulting layer, not the other way around.

24/7 embedded operator model for home care agencies where two founders personally run agentic hiring/scheduling as a managed service — the product and the delivery are the same thing.

The offer IS an AI agent stack trained on 12,000 proprietary petitions, with attorneys as QA — a genuine product-with-legal-wrapper, not consulting with AI in the copy.

The diagnostic itself is an AI product — agents run 280 parallel interviews in 9 days that no consultant could; the consulting work wraps around the AI, not the other way around.

Ingest regulatory proposals across jurisdictions → AI assesses business impact → delivers strategic advocacy guidance.

"AI-native law firm. AI does 80%, elite lawyers do the final 20%. $2bn in deals, 4-hr average turnaround." Tests D2 (builds AND runs the legal capability), D5 (vertical wedge), D6 (concrete outcome numbers published). Expect 4s and 5s on several dimensions — a deep-generalist or wedge shape.