Question 1

What is LeapFrog?

Accepted Answer

LeapFrog is a multi-cloud AI platform that sees, diagnoses, and fixes itself. We detect every operational issue, security alert, cost inefficiency, and high-ROI cloud migration opportunity and provide an immediately executable plan — with the option to auto-apply or stage through approvals. Kermit, our unified AI DevOps engineer, operates inside your cloud accounts with least-privilege access, so you retain ownership, visibility, and control.

Question 2

How does LeapFrog work?

Accepted Answer

You describe the desired outcome in natural language (for example: 'deploy a production-ready, autoscaled API with blue/green releases' or 'reduce EC2 spend by 30% without affecting SLOs'). Kermit compiles that intent into an executable plan, generates Terraform-like changes, simulates impact, and executes with policy guardrails. Telemetry (logs, metrics, traces, and cloud events) and config graphs are continuously analyzed to propose optimizations and fixes. Every change is diffed, signed, and auditable.

Question 3

What does Kermit actually do end-to-end?

Accepted Answer

Provisioning (networks, clusters, services, CI/CD), configuration, observability wiring, cost and performance optimization, incident detection and root-cause analysis, automated remediation with tests and rollback, compliance checks, and cross-cloud migrations. Kermit maintains an always-current model of your estate (resources, dependencies, policies) and reasons over it to recommend and ship changes.

Question 4

Where does LeapFrog run and who owns the infrastructure?

Accepted Answer

You do. LeapFrog connects to your AWS, GCP, Azure (and 500+ other cloud) accounts using scoped roles and service principals. Execution occurs in your environment; artifacts (plans, diffs, logs) are stored in your chosen locations. We prioritize least-privilege, short-lived credentials, and tamper-evident audit trails.

Question 5

How do we integrate LeapFrog with our stack?

Accepted Answer

Connect cloud accounts (read-only or with change-apply permissions), point to your observability sources (CloudWatch, Stackdriver, Azure Monitor, OpenTelemetry, Prometheus), link VCS for policy and review workflows, and define guardrails (approval policies, maintenance windows, environments). Kermit discovers resources, builds the graph, and begins proposing safe changes.

Question 6

Does LeapFrog replace Terraform, Pulumi, or Helm?

Accepted Answer

No - we're compatible. Kermit can generate and operate standard artifacts (e.g., Terraform-like modules, Helm charts) and commit them to your repos with PRs for review, or apply them directly through approved pipelines. If you already have IaC, Kermit reads it, detects drift, and proposes consistent changes.

Question 7

What observability does LeapFrog require?

Accepted Answer

We ingest signals from your existing stack (OpenTelemetry, Prometheus, cloud provider metrics/logs, tracing backends) and - if missing - Kermit can instrument default dashboards, alerts, and SLOs. Signal quality gates ensure we only auto-remediate when confidence thresholds and rollback conditions are satisfied.

Question 8

How safe is auto-remediation?

Accepted Answer

Every remediation is a plan with pre-checks, blast-radius limits, canaries, and automated rollback. You can choose 'proposal-only', 'require approval', or 'fully automated' per environment, service, or policy. Kermit includes synthetic tests and post-change verification before declaring incidents resolved.

Question 9

How do migrations work?

Accepted Answer

Kermit scans your estate, maps dependencies, and produces a migration workbook: target architecture, IaC for the destination, data movement strategy, and a cutover plan with verification and rollback. Dry-runs validate capacity, performance, and cost. Execution can be phased by service to minimize risk.

Question 10

What about compliance and audits?

Accepted Answer

We maintain tamper-evident logs of every plan, approval, and apply. Policies encode separation of duties and change windows. Evidence packs (who changed what, when, and why) are exportable for SOC 2/ISO 27001 audits. Data residency controls and logging destinations are configurable.

Question 11

How is pricing structured?

Accepted Answer

Pricing aligns with value and scale: a platform fee plus usage-based tiers (environments, managed services, and optional automation features). We support proof-of-value pilots that start in read-only mode and expand to change-apply once policies are in place.

Question 12

What are the limits and best practices?

Accepted Answer

Start with read-only discovery to build the estate graph. Define guardrails (who approves where), attach SLOs to critical services, and enable auto-remediation first for low-risk classes (e.g., autoscaling, cache tuning). Graduate to higher-impact optimizations once verification tests are in place.

Question 13

How do we get started?

Accepted Answer

Book a demo. We'll connect to a sandbox or read-only roles, generate a current-state map, and show immediate wins (e.g., cost reductions, reliability fixes). From there, enable change-apply with your approval policies and begin automating day-to-day DevOps work with Kermit.

Multi-cloud AI platform that sees, diagnoses, and fixes itself

Comprehensive Cloud Infrastructure Management

Automated Task Management

Cloud Migrations Made Simple

Multi-Cloud Data Warehousing

AI-Powered Assistant

Transform Your Infrastructure Operations

With LeapFrog

Without LeapFrog

Works with all major cloud providers

Make infrastructure your winning advantage