BigPanda
Also known as: bigpanda.io, BigPanda Biggy, Biggy AI, BigPanda Agentic ITOps Platform, BigPanda AI Incident Assistant, BigPanda AI Detection and Response
BigPanda is an agentic IT operations platform whose Biggy agents triage alerts, investigate incidents across observability tools, coordinate major incidents in Slack, Teams and ServiceNow, and assess change risk. A reasoning core composes integration agents, built in skills and custom agents, with adaptive memory of past fixes, approval before any write action and a full audit log. It runs on AWS with separate US and EU instances, offers an API and an MCP server, and sells credit plans by quote.
BigPanda sells an agentic IT operations platform to the teams that keep enterprise services running, with customers such as UBS, IHG Hotels and Resorts, the London Stock Exchange, Labcorp, PlayStation and Bread Financial. It began with AIOps event correlation, bought the SRE startup Velocity in November 2025, and now puts AI agents on L1 operations, major incident response and change risk, with Biggy as the agent responders talk to.
Biggy works in a web app, Slack, Microsoft Teams, email and the ServiceNow Now Assist panel. A reasoning core sizes each request as a quick answer, a focused lookup or an intensive investigation and composes the integration agents, actions and skills it needs, sending several agents after competing hypotheses at once when an outage is unclear. Thirteen built in skills such as Generate RCA, Executive Summary and 5 Whys Analysis sit beside skills a team writes in plain language, custom agents with their own persona, tools and scope, and workflows drawn on a visual canvas with branch, approval and wait steps.
Triggers start work on a schedule or an event, ITSM agents watch ServiceNow for matching records, and Biggy can join a Teams bridge call, post major incident updates and draft the post incident report. Underneath, BigPanda Core correlates alerts from tools such as Datadog, Dynatrace, Splunk, New Relic and CloudWatch, the IT Knowledge Graph joins tools, tickets and team knowledge, and Adaptive Memory keeps facts, proven fixes, learned mistakes and past episodes, each tied to its source and open to admin review.
Any action that changes a system, such as opening a ticket, paging on call or resolving an incident, pauses for approval in the web app, Slack or Teams, where the approver can edit each field, and unattended runs stop at write steps unless an admin allows them. An audit log records what Biggy did for each user, roles decide who can build and who can call each agent, and sign in runs through single sign on with Okta, OneLogin or PingOne.
The company has passed a SOC 2 Type II audit every year since 2017 and runs on AWS with separate US and EU instances, including a fully EU resident Biggy, and the model can run on OpenAI through BigPanda or on Azure OpenAI in BigPanda's tenant or the customer's own. Developers reach it through the Biggy API, an MCP server for editors and chat apps, and an MCP client for tools without a native integration. Pricing is by quote, with one credit pool across every product and tiered plans from 20,000 credits on one to three year terms.
Its natural buyer is a large enterprise with a busy operations center, a ServiceNow estate and many monitoring tools, wanting agents to take on L1 triage, run major incident coordination and screen change risk while engineers keep the final say on every write. It is a weaker match for a small engineering team that wants a self serve trial, published prices or agents that work through a browser.
Vendor details
Canonical URL
https://www.bigpanda.io
Category
SRE / DevOps agent
Subcategory
Agentic IT operations platform whose Biggy agents triage alerts, investigate and coordinate incidents, automate L1 operations and assess change risk on top of AIOps event correlation
Funding status
Venture backed private company; investors include Advent International, Insight Partners, Battery Ventures and Sequoia. Bought the SRE startup Velocity in November 2025.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
BigPanda Core takes alerts, changes and topology from about 50 listed tools including Datadog, Dynatrace, Splunk, New Relic, AppDynamics, Prometheus, Grafana, Amazon CloudWatch, Azure Monitor, SolarWinds, Zabbix, ThousandEyes, Jenkins, AWS CloudTrail and VMware vCenter; ServiceNow (native app and Now Assist agent) and Jira Service Management; Slack, Microsoft Teams, email, PagerDuty, OpsGenie, xMatters, Everbridge and Grafana OnCall; Okta, OneLogin and PingOne for SSO; any MCP server or custom REST API.
In practice
A bank's operations center gets a storm of alerts from Dynatrace and Splunk overnight. BigPanda correlates them into one incident, Biggy adds the likely root cause and a recommended action to the ServiceNow ticket, and the L1 analyst approves the page to the database team from the Slack card.
During a major outage the incident commander pastes the Teams bridge link into the incident template. Biggy joins the meeting chat, posts scheduled status updates until the incident is mitigated, and drafts the post incident report from the timeline and transcript.
A change advisory board reviews Biggy's risk assessment for each pending change, built from past incidents on the same services, approves the low risk changes faster and holds the risky ones for a closer look.
Sources & related URLs
Related / legacy domains
Research sources
Agentic Index coverage score
12.0 / 14 capabilities · 86%
| Integrations & Tool Calling | Full |
|---|---|
|
BigPanda Core takes alerts, change events and topology from about 50 listed tools, among them Datadog, Dynatrace, Splunk, New Relic, AppDynamics, Prometheus, CloudWatch, Azure Monitor, Jenkins, CloudTrail and VMware vCenter. Biggy's agents query those tools live, create and update ServiceNow and Jira records, retrieve on call from PagerDuty, OpsGenie or xMatters, and reach in house systems through a custom REST connection or any MCP server. Sourcebigpanda.io/integrationsread 2026-11-19 |
|
| Workflow Orchestration | Full |
|
A reasoning core reads each request, sizes it as a quick answer, a focused lookup or an intensive investigation, and composes the integration agents, actions and skills it needs, while a swarm investigation sends several agents after competing hypotheses at once. Procedures that must run the same way every time go on a visual workflow canvas with action, prompt, branch, approval, wait, abort and output steps, versioned and launched by trigger or command. Sourcedocs.bigpanda.io/en/improved-ai-incident-assistant-and-prevention-web-appread 2026-11-19 |
|
| Knowledge Grounding & RAG | Full |
|
The IT Knowledge Graph joins monitoring tools, tickets, change records, topology and the tribal knowledge held in past incidents, so the agents answer from the company's own environment, and Biggy pulls the knowledge base and similar incidents into each investigation. Admins choose which data sources Biggy may read, and Knowledge Corrections let a team fix what Biggy has learned so later answers follow the corrected version. Sourcebigpanda.ioread 2026-11-19 |
|
| Human Oversight & Guardrails | Full |
|
Any write action, such as changing a ticket, paging on call, resolving a BigPanda incident or calling an MCP tool that changes data, pauses for a person's approval on a card in the web app, Slack or Teams, where the approver can edit values field by field. Destructive steps need a second confirmation, unattended or scheduled runs stop at write actions unless an admin allows them, and a safety classifier and configurable guardrails screen every request. Sourcedocs.bigpanda.io/en/improved-ai-incident-assistant-and-prevention-web-appread 2026-11-19 |
|
| Security, Identity & Governance | Full |
|
The company has passed an independent SOC 2 Type II audit every year since 2017, and its program follows an ISO 27002 based framework on AWS. Sign in runs through single sign on with providers such as Okta, OneLogin and PingOne, roles decide who may create agents and who may call each one, service keys carry scoped permissions, and data is encrypted with 256 bit AES at rest, TLS 1.2 in transit and keys held in AWS KMS and CloudHSM. Sourcedocs.bigpanda.io/en/securityread 2026-11-19 |
|
| Observability & Auditability | Full |
|
An audit log records what Biggy did on each user's behalf, actions taken with service keys appear in the activity log, and a settings history tracks every change to skills, workflows, actions and guardrails with diffs and one click restore. The AI Learning page shows feedback on Biggy's answers and agent runs, the Biggy API exports full usage data, and Unified Analytics reports on incident volumes and response across the estate. Sourcedocs.bigpanda.io/en/bigpanda-biggy-ai-release-notesread 2026-11-19 |
|
| Memory & State Persistence | Full |
|
Adaptive Memory keeps four kinds of knowledge between conversations: facts, proven approaches, learned mistakes and episodes, each linked to its source with a confidence level and a count of how often it has held up. A background process consolidates memory between sessions, past incidents and proven fixes carry into new conversations, Slack and Teams threads are remembered, and an admin page lets the team review what Biggy has kept. Sourcedocs.bigpanda.io/en/improved-ai-incident-assistant-and-prevention-web-appread 2026-11-19 |
|
| Deployment & Data Residency | Full |
|
Customers in the United States use one hosted instance and customers in the UK and EU another, and Biggy EU runs as a fully EU resident instance that keeps data in the region end to end. The model can run on OpenAI through BigPanda or on Azure OpenAI, either hosted by BigPanda or in the customer's own tenant, and relay clients connect tools inside a customer's network. There is no self hosted edition. Sourcedocs.bigpanda.io/en/securityread 2026-11-19 |
|
| Prebuilt Agents / Templates / Packs | Full |
|
Thirteen built in skills ship ready, among them Generate RCA, Executive Summary and 5 Whys Analysis, alongside integration agents for ServiceNow, Datadog, Splunk and Dynatrace and stock actions for channel summaries, incident channels, ITSM records and on call lookups. Major incident templates, Suggested Actions, Change Risk Management and Problem Management come as ready modules, and skills and templates can be exported and shared between teams. Sourcedocs.bigpanda.io/en/improved-ai-incident-assistant-and-prevention-web-appread 2026-11-19 |
|
| Triggers & Channel Coverage | Full |
|
Work starts in many places: chat in the web app, Slack or Microsoft Teams, the ServiceNow Now Assist panel, schedules and events set as triggers, and ITSM agents that watch ServiceNow for matching records and analyze them on their own. Biggy can join a Teams bridge call from a link in a major incident template, post updates back to the originating Slack thread, and send each incident's updates by email to approved domains. Sourcedocs.bigpanda.io/en/bigpanda-biggy-ai-release-notesread 2026-11-19 |
|
| Model Flexibility & Routing | Partial |
|
Biggy runs on OpenAI models by default, through a service with zero data retention, and an organization can switch to Azure OpenAI, hosted either by BigPanda or in its own Azure tenant with its own model deployments. Changing the provider needs BigPanda's approval, the choice stays within OpenAI models, and there is no per task picker or routing across vendors beyond Biggy sizing the effort each request gets. Sourcedocs.bigpanda.io/en/manage-ai-incident-assistant-configurationread 2026-11-19 |
|
| APIs / SDKs / MCP Extensibility | Full |
|
The Biggy API manages keys, users, roles and team groupings, exports usage, and exposes action plans, transcripts and A2A endpoints, and a remote MCP server lets editors and chat apps query incidents, changes, on call status and the knowledge base, signing in through OAuth 2.1. Biggy also works as an MCP client for a team's own servers, its action plans can be called as remote tools, and the REST API takes alerts from any source. Sourceapi-docs.bigpanda.io/mcpread 2026-11-19 |
|
| Testing, Debugging & Optimization | Partial |
|
Custom agents can be tried in an Agent Preview panel before they are published, and workflows support dry runs and test runs. Users rate Biggy's answers and ITSM agent runs, and the AI Learning page shows that feedback over time. There is no test set, scoring against past incidents or side by side comparison of agent versions, so quality is judged from previews and live feedback. Sourcedocs.bigpanda.io/en/improved-ai-incident-assistant-and-prevention-web-appread 2026-11-19 |
|
| Browser / Computer-use | Not documented |
|
Biggy reaches systems through integrations, its own API, MCP servers and custom REST connections, and acts by calling those tools. None of its agents drives a browser or a desktop session to work through a console, a vendor portal or an internal page that lacks a connector, so a system without an API or MCP route stays outside what Biggy can operate. Sourcedocs.bigpanda.io/en/integrate-with-ai-incident-assistant-and-preventionread 2026-11-19 |
|
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
Quote only; credit plans from 20,000 credits on one to three year terms
Credits from one pool shared across all BigPanda products, on one to three year commitments
Included quota
Tiered credit plans start at 20,000 credits
What is public
BigPanda prices by quote. Every product draws on one pool of credits, and tiered credit plans start at 20,000 credits.
Billing mechanics
Contracts commit for one to three years, and the single credit pool is spent across all BigPanda products rather than a license per module.
Cost watchouts
Commitments run one to three years, and rollout help comes as separate professional services packages.
Variable cost rationale
Every product draws on one committed credit pool, so heavier use or added products spend the commitment faster.
Additional watchouts
No dollar price per credit is shown, so the cost of a 20,000 credit plan needs a quote.
Sales call required
Yes, required for paid access
Free / trial
No free tier or self serve trial; proof of value assessments offered
Commercial notes
A proof of value assessment is offered before purchase, and BigPanda is also sold through AWS Marketplace.
Key ambiguities
How many credits each product or Biggy action consumes is not stated.
Cancellation / refund
One to three year commitments.
Support SLA / resale
Professional services packages are listed alongside the platform.
Missing data
BigPanda gives no price per credit, no plan names and no trial terms.
Related vendors
- AlertD — AI agents for AWS operations that run read only in the customer's…
- Anyshift — AI SRE built on a versioned knowledge graph of infrastructure, apps…
- Avesha — Obliq, Avesha's autonomous AI SRE for Kubernetes and agentic…
- Better Stack — Better Stack offers incident management with built in on call,…
- Bluebricks — Context and control layer that lets AI agents operate cloud…
- Cased — AI workflows for infrastructure and platform engineers: default and…