Back to vendors
C

Coralogix

Also known as: Coralogix Olly, Olly AI, Coralogix AI Center

Visit site
Entry priceFree monthly AI Unit allowances, paid usage by model and planFull pricing detail

Coralogix Olly investigates problems using logs, metrics, traces, and other data in a team's observability account. It can run saved prompts on a schedule, watch selected dashboards, and send findings to configured notification channels. Teams on eligible plans can also trigger investigations when a matching case opens. Coralogix includes an API for calling Olly from other tools and an AI Center for monitoring and evaluating AI applications the buyer operates.

Coralogix gives engineering and security teams a way to investigate incidents across logs, metrics, and traces, and sells observability and SIEM on one platform. Olly, its team of AI agents, can formulate queries, follow evidence across telemetry sources, and return structured findings inside Coralogix.

On security work it hunts across identity, endpoint, cloud and application telemetry from sources such as Okta, CrowdStrike Falcon, AWS CloudTrail and Wiz, maps findings to MITRE ATT&CK and proposes containment such as isolating a host, revoking a session or blocking an IP, which runs only after approval. It can work with a selected GitHub repository and with supported external connectors when those integrations are enabled. A team can supply instructions and investigation skills for recurring work. Olly retains selected facts and preferences for individual users across chats, and users can see and manage what it saves.

A scheduled task runs a saved prompt at a chosen time and keeps each run's conversation in its history. A user can also configure a dashboard watch and approve the watch before it starts. Results can go to supported notification destinations such as Slack or Microsoft Teams. Automatic case analysis is available to eligible paid customers in early access. When a case matches their criteria, Olly can investigate the event, correlate the underlying telemetry, form and test hypotheses, and post its findings directly to the case. The case's notification rules can pass those findings to the people handling the incident.

Teams can use the Olly REST API to start investigations and retrieve the artifacts they produce. Each user chooses GPT, Claude or Gemini and how much reasoning effort Olly spends, with access depending on region. Coralogix aligns processing with selected hosting regions for eligible model and location combinations. AI Center, in the same platform, applies configured policies to streams from the buyer's AI applications.

Separate AI Units measure the model and token usage for Olly and AI Center. The free plan includes 10 units per user per month and a shared team limit of 100, and a 14 day trial opens every feature with no credit card. Paid plans provide additional capacity, with exact charges and limits depending on the account. Coralogix holds SOC 2, ISO 27001, PCI DSS and FedRAMP Moderate and supports HIPAA.

More than 4,000 companies use Coralogix, including more than 20 of the Fortune 500 that each send over 100 TB a day, among them Adobe, Tradeweb and Bank Jago, and Gartner named it a Leader in its 2026 Magic Quadrant for Observability Platforms. Simpplr's engineers cut some troubleshooting from hours to minutes with Olly working over the logs and telemetry already in Coralogix. An on call team can set up a dashboard watch to notice a recurring latency change, review the queries and findings, and send the result to the people responsible for the service.

A platform team working on its own AI application can use AI Center to track outputs against a selected policy. The team can inspect those results beside application telemetry and follow up when a workflow behaves differently from what it expects. This makes the product relevant both to the on call responders investigating an incident and to the application owners reviewing the quality of their automated systems. A buyer can configure permissions and access independently for those jobs.

Vendor details

Canonical URL

https://coralogix.com

Category

SRE / DevOps agent

Subcategory

Observability and incident investigation agents

Funding status

Coralogix is a privately held, venture backed software company that raised $200 million to scale its observability platform.

Company status

independent

Use cases & customers

Primary use cases

Investigate application symptoms against logs, metrics, and tracesRun dashboard watches and recurring operational checksEvaluate and monitor AI applications the buyer runs

Target customers

Site reliability and DevOps teams investigating telemetrySecurity operations teams responding to configured casesPlatform teams evaluating their own AI applications

Deployment options

Managed regional Coralogix deploymentsRegional model availability and customer storage integrations

Integrations

Olly can reach connected GitHub repositories, configured external tools in public beta, security sources such as Okta, CrowdStrike Falcon, AWS CloudTrail and Wiz, and notification channels such as Slack and Microsoft Teams. A REST API starts investigations and returns their artifacts.

In practice

An on call team schedules a dashboard watch for error rates and sends findings from each relevant run to its operations channel.

An eligible paid team enables automatic case analysis and reviews root cause findings posted to cases that match its conditions.

An AI platform team streams traces from an assistant it built into AI Center and reviews policy scores for application outputs.

Agentic Index coverage score

13.0 / 14 capabilities · 93%

Integrations & Tool Calling Full

A configured agent task can run a query over buyer telemetry and send an output to a connected external notification channel. Supported connectors can also bring contextual data from external systems, subject to availability and access. A workflow can write a finding to a case in an eligible deployment. Destinations and allowed actions depend on the configuration.

Sourcecoralogix.com/docs/user-guides/olly/dashboard-watchread 2026-11-20

Workflow Orchestration Full

An automatic investigation can start with a matching case, gather telemetry, identify a possible cause, and record a finding. The service can send the result to destinations chosen for the case. The case feature is early access for eligible paid teams. The core agent can also coordinate evidence gathering and query refinements while running scheduled or prompted investigations.

Sourcecoralogix.com/docs/user-guides/olly/automatic-case-analysisread 2026-11-20

Knowledge Grounding & RAG Full

Olly uses the team's maintained Coralogix data and can work across its stored logs, metrics, traces and alerts, with queries limited to the resources and context the caller can access. When enabled, it adds selected repository material from a connected GitHub account to investigate code. That searchable telemetry base serves every investigation as the underlying data changes.

Sourcecoralogix.com/docs/user-guides/ollyread 2026-11-20

Human Oversight & Guardrails Full

Olly can propose a dashboard watch and present its configuration for people to review before activation, and the operator can revise what the watch checks and where results go. On security cases, containment actions such as isolating a host, revoking a session or blocking an IP wait for approval. Teams also control connector permissions and authorized destinations.

Sourcecoralogix.com/docs/user-guides/olly/dashboard-watchread 2026-11-20

Security, Identity & Governance Full

Coralogix holds SOC 2, ISO 27001, PCI DSS and FedRAMP Moderate. Tenant data is separated by customer keys and region, code and prompts are encrypted and handled under restrictions, and the agent inherits platform access permissions over telemetry and allowed connected accounts. Administrators configure access and can limit network availability.

Sourcecoralogix.com/docs/user-guides/olly/data-privacyread 2026-11-20

Observability & Auditability Full

A scheduled task keeps dedicated chat and run history, and the REST API serves interactions and structured artifacts. An operator can use the output and query results to inspect an investigation. Buyers using AI Center can also view policy results associated with traces for their own applications. Usage views provide additional cost detail by activity.

Sourcecoralogix.com/docs/developer-portal/apis/ollyread 2026-11-20

Memory & State Persistence Full

The product keeps selected facts and preferences from interactions and applies them during later chats. A memory can describe system terminology or a service that a particular user investigates often. The user's controls allow them to manage these stored items. Memories are scoped to the individual and aren't shared with other users of the Coralogix team.

Sourcecoralogix.com/docs/user-guides/olly/memoriesread 2026-11-20

Deployment & Data Residency Full

Coralogix runs in selected hosting regions, and model processing aligns with the customer's region, with the service handling inference for each provider set per region and Gemini unavailable in US regions. Model providers run on dedicated infrastructure Coralogix manages, apart from storage the buyer owns, and a FedRAMP Moderate environment serves US government work.

Sourcecoralogix.com/docs/user-guides/olly/data-privacyread 2026-11-20

Prebuilt Agents / Templates / Packs Partial

The agent loads matching supplied skills as well as team and personal skills, where a team sets investigation steps, terminology and response patterns. Modes such as dashboard watches and automatic case triage come with the product, but most of these workflows depend on the buyer configuring its own data, prompts and destinations.

Sourcecoralogix.com/docs/user-guides/olly/skillsread 2026-11-20

Triggers & Channel Coverage Full

A user can schedule a saved task for a one time or recurring run, each run creating its own chat in the task's history, and a dashboard watch uses the same scheduling. Automatic analysis when a case opens is in early access for eligible paid accounts. The REST interface also lets an outside system start an investigation.

Sourcecoralogix.com/docs/user-guides/olly/scheduled-tasksread 2026-11-20

Model Flexibility & Routing Full

Each user chooses GPT, Claude or Gemini for a chat, and how much reasoning effort Olly spends, weighing complexity, cost and speed. The choice persists for later chats until changed, and availability follows the account's region and provider processing terms. The API offers the same model choices as the product.

Sourcecoralogix.com/docs/user-guides/olly/model-selectionread 2026-11-20

APIs / SDKs / MCP Extensibility Full

An outside client can start an Olly investigation, send a prompt and retrieve results through the REST API, which also covers chat and interaction resources, structured artifacts, schedules and configuration. Teams can add Olly to an existing automation or feed its outputs to another agent within the caller's permissions, and each call returns a defined set of objects.

Sourcecoralogix.com/docs/developer-portal/apis/ollyread 2026-11-20

Testing, Debugging & Optimization Full

AI Center can apply policies to spans from an AI application the customer runs. A buyer configures the policies and evaluators, and the system scores the incoming work for the corresponding behavior. The platform supports applying those evaluators to development and production traces, so an owner can inspect the score and see where an application changes.

Sourcecoralogix.com/docs/user-guides/ai/evaluationsread 2026-11-20

Browser / Computer-use Partial

When an administrator allows it in the team's network settings, Olly can fetch a public URL it learns from a prompt or a permitted tool result and fold the page contents into its investigation. The agent reads identified webpages, with no interactive browser or desktop session to drive.

Sourcecoralogix.com/docs/user-guides/olly/data-privacyread 2026-11-20

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Pricing

Free monthly AI Unit allowances, paid usage by model and plan

Per-model input and output token consumption measured in AI Units

Free tier

What is public

The free allowance is 10 AI Units per user and 100 across the team each month, resetting on the first of the month. A 14 day trial with no credit card opens every feature.

Billing mechanics

Olly and AI Center use an AI Unit pool separate from telemetry units. AI consumption depends on input tokens, output tokens, and which model was used.

Variable cost rationale

AI Units meter input and output tokens by model, so heavier investigations, more reasoning effort and pricier models all raise spend, on top of telemetry billing.

Additional watchouts

Free features stop for the user or team after its applicable AI allowance is exhausted. Automatic case analysis requires an eligible paid account.

Sales call required

Mixed (some tiers require a call)

Free / trial

Free plan with 10 AI Units per user and 100 per team each month

Missing data

Paid plan commitments, included AI Unit capacity and effective rates depend on the account agreement.

Agentic Index verified 2026-11-20

Alternatives to Coralogix

Other products that do the work Coralogix does, and what each one is.

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.