RXed AI News

AI news with a spine.
@RXed_EU
SCORE 7/1018/20 elements Audited 2026-07-29 · RXed table v1.0

Sierra

“AI agent platform for customer experience” — the vendor’s own words

Outcome pricing took Sierra from $100M to $200M ARR in a year, but the receipt is a black box: no rate card, a ~€130K+ platform floor, and roughly €1.30 per resolved conversation negotiated behind an NDA.

Enterprise platformAutomation & AgentsChatSalesFreemium
Vendor
Sierra · sierra.ai
Origin
US — San Francisco
Pricing
Annual platform minimum ~$150,000-750,000+/yr · Implementation/setup ~$50,000-200,000 · Per-resolution fee ~$1.00-2.50 (reported ~$1.50 typical)
Users (official only)
Serves over 40% of the Fortune 50 and one in three of the world's largest banks; 30%+ of customers have annual revenue over $10B, 50%+ over $1B. (source, 2026-07-14)
Annual platform minimum~$150,000-750,000+/yrcustom, outcome-based; scales with volume/channels/complexity
Implementation/setup~$50,000-200,000one-time, 4-10 week typical onboarding, up to 7 months for large rollouts
Per-resolution fee~$1.00-2.50 (reported ~$1.50 typical)negotiated per contract; unresolved/escalated conversations typically don't trigger a charge

Fully custom, sales-led, no public rate card. No included helpdesk/ticketing — runs as an AI layer above existing CX infrastructure (Zendesk, Salesforce Service Cloud, etc.).

checked 2026-07-29 · vendor pricing page

Element scores

Reactive
Retrieval & Memory
Orchestration
Validation
Models
Primitives
Pr8
Prompts
Em6
Embeddings
Cx8.5
Context
Tr8.5
Tracing
Lg8.5
LLM
Compositions
Fc8.5
Function calling
Vx5.5
Vector store
Rg8.5
RAG
Gr8
Guardrails
Mm6
Multimodal
Deployment
Ag9.5
Agents
Ft
Fine-tuning
Fw9
Frameworks & harnesses
Ev9.5
Evaluations
Sm6
Small models
Emerging
Ma6.5
Multi-agent
Sy
Synthetic data
Pc7
Protocols
In4
Interpretability
Th5.5
Thinking models
Tap or hover any element to see why it got that score.

Strengths

Agent Data Platform gives it a real memory layer most competitors lack, the model-agnostic "constellation" approach avoids single-vendor lock-in, and Sierra Research built τ-bench — the benchmark the rest of the industry now evaluates agents against. Case studies are backed by real speed-to-live numbers: Nordstrom's voice agent in five weeks, Cigna cutting patient-auth time 80% in eight weeks.

Honest dings

Zero pricing transparency — no rate card, no self-serve, a six-figure floor before a single resolution is billed. G2 reviewers flag context loss in long conversations and generic responses; the December 2025 Gap.com incident showed guardrails are only as good as their configuration. Multi-model orchestration adds real latency risk on live voice.

Best for: Large consumer or B2B enterprises (Fortune 500-scale, high ticket volume) that can absorb a six-figure sales cycle and want an AI layer bolted onto existing CX infrastructure rather than a full helpdesk replacement.
Visit sierra.ai Prices and details change — this passport is re-verified at least quarterly.

Sources

Every audit lists the research it rests on — transparency and traceability are the product. Tools evolve: each audit is a snapshot of its audit date, and re-audits supersede older versions (kept below for reference).