Skip to content

Traceloop

★ New
assess
Observability vendor Commercial (SaaS) + Apache-2.0 (open-source SDK) freemium

At a Glance

Israeli AI observability startup (YC W23) behind OpenLLMetry, acquired by ServiceNow in March 2026 for $60–80M; its technology is being integrated into ServiceNow's AI Control Tower enterprise governance platform.

Type
vendor
Pricing
freemium
License
Commercial
Adoption fit
small, medium, enterprise
Top alternatives

What It Does

Traceloop is the company behind OpenLLMetry, an open-source LLM observability library. The company also operated a commercial SaaS platform offering hosted LLM tracing, evaluation dashboards, and prompt management layered on top of OpenLLMetry’s open-source core. Its core differentiation was OTel-native architecture — routing LLM telemetry into existing APM stacks rather than requiring a separate observability platform.

In March 2026, Traceloop was acquired by ServiceNow for an estimated $60–80M. ServiceNow’s stated intent is to integrate Traceloop’s technology into its AI Control Tower product for enterprise AI governance — automated agent evaluation, behavior tracking, and observability within ServiceNow’s enterprise IT platform.

Key Features

  • Hosted tracing UI with span and token-level drill-down for LLM calls
  • Prompt versioning and management linked to trace data
  • Integration with 23+ observability backends via OTLP (Datadog, Grafana, Honeycomb, New Relic, SigNoz, etc.)
  • Open-source SDK (OpenLLMetry) under Apache-2.0 with no SaaS dependency
  • Contributions to OpenTelemetry GenAI Semantic Conventions SIG (now part of the official OTel spec)
  • 7,000+ GitHub stars across OpenLLMetry repositories; $6.6M seed funding (Sorenson Capital, Y Combinator, Samsung NEXT)

Use Cases

  • Use case 1: Teams wanting managed LLM observability without operating their own OTel pipeline — Traceloop’s SaaS platform as backend
  • Use case 2: Enterprise teams evaluating AI governance tooling that will be embedded in ServiceNow’s AI Control Tower post-acquisition
  • Use case 3: Organizations contributing to or tracking the OpenTelemetry GenAI Semantic Conventions standards process

Adoption Level Analysis

Small teams (<20 engineers): Viable via the free tier of the Traceloop SaaS. Post-acquisition, SaaS availability and pricing need confirmation; teams starting fresh should evaluate Langfuse (MIT, self-hosted) as a lower-risk baseline.

Medium orgs (20–200 engineers): The Traceloop platform adds value over raw OpenLLMetry when teams need hosted trace storage, prompt management, and evaluation UI without operating Langfuse or similar self-hosted stacks. Post-acquisition roadmap clarity is needed before committing.

Enterprise (200+ engineers): ServiceNow’s acquisition targets this segment specifically. If already a ServiceNow customer, AI Control Tower integration may make sense. Non-ServiceNow enterprises should treat this as a vendor risk flag — the commercial product roadmap is now dictated by ServiceNow’s enterprise strategy, not by the independent AI observability market.

Alternatives

AlternativeKey DifferencePrefer when…
OpenLLMetry (standalone)Apache-2.0 SDK; no SaaS dependencyYou want Traceloop’s OTel instrumentation without commercial platform lock-in
LangfuseMIT; self-hosted; full-stack (tracing + eval + prompt mgmt + datasets); 24k+ starsYou want an all-in-one observability platform that is independent and actively maintained
LangSmithProprietary; deepest LangChain agent debugging; $39/user/monthLangChain-heavy stack requiring first-class agent chain visualization
Arize PhoenixOTel-native; strong evaluation toolkit; local-first devNeed evaluation co-located with tracing; no APM vendor preference

Evidence & Sources

Notes & Caveats

  • Acquisition risk is high: The March 2026 ServiceNow acquisition is the dominant risk factor for any team evaluating Traceloop’s commercial platform. ServiceNow has a history of integrating acquired products into its enterprise suite, often at enterprise pricing tiers that exclude smaller teams. Monitor the AI Control Tower roadmap closely.
  • Open-source commitment: The team has committed to keeping OpenLLMetry Apache-2.0 and continuing OTel contributions. The open-source library is the safer long-term bet than the commercial platform.
  • YC W23 + small team: At acquisition, Traceloop had an ~8-person team reporting $1.2M ARR (2024). Engineering depth was heavily concentrated in the founding team; ServiceNow integration may dilute that focus on the open-source project.
  • Prior silent telemetry: Versions of OpenLLMetry before v0.49.2 collected usage telemetry without explicit user consent. The behavior was fixed after community pressure. This is relevant to due diligence for earlier adopters.

Related