Prerequisites

Here is the list of prerequisites to start your journey with the Flashback platform. We also recommend you to read the explanation about the platform :

• Supported Browsers

• Vendors' Cloud Accounts

• CLI and SDK Tools

• Supported Platform Services

• Future Platform Capabilities

---

Supported Browsers

Latest stable Chrome, Firefox, or Safari (ensure JavaScript and cookies are enabled).

---

Vendors' Cloud Accounts

Flashback is a solution that connects you to your existing provider, such as Amazon Web Services, Google Cloud Platform, Microsoft Azure, OpenAI, Anthropic AI, StorJ or Filecoin's Akave.

Provider	Account Creation
AWS	Create account
GCP	Signup
Azure	Build in the cloud with an Azure account
StorJ	Signup
Akave	Try Akave Cloud for free today
OpenAI	Welcome
Anthropic AI	Start With

We recommend you to visit the documentation of your favorite cloud providers to learn how to set up efficiently your resource enabler (storage, AI LLMs, and more).

---

CLI and SDK Tools

You can use native CLI and SDK of providers to help you to configure your resources:

• Preferred post-2020 versions of programming language version for SDK integrations.

Provider	Command Line Interface	Software Development Kit
AWS	What is the AWS Command Line Interface?	Toolbox
GCP	Google Cloud Command Line Interface (gcloud CLI)	Download and install Google Cloud SDK
Azure	Azure Command-Line Interface (CLI) documentation	SDKs and tools for managing and interacting with Azure services
OpenAI	Not found	Libraries
Anthropic AI	CLI Reference	Client SDKs

For cloud storage, you can also use the S3-compatible APIs (non-exhaustive list):

Provider	S3-compatible APIs
StorJ	S3 Compatibility
Akave	Akave O3
MinIO (on-prem)	S3 Compatible Storage

Basic Cloud-Storage Knowledge

• Understanding buckets, objects, ACLs, IAM roles, signed URLs. • Familiarity with uploading/downloading via CLI or REST APIs.

---

Supported Platform Services

Object Storage

Find and understand the different object storage solutions from the providers.

Provider	Object Storage Creation
AWS	Getting started with Amazon S3
GCP	Create a bucket
Azure	Create an Azure storage account
StorJ	Create a bucket
Akave (Filecoin)	Create a bucket

We recommend you to visit the documentation of your favorite cloud providers supporting S3-compatible APIs to learn how to set up efficiently your storage enabler.

AI LLMs

Find and understand the different AI LLM solutions from the providers.

Provider	Start with AI LLMs
AWS	Getting started with Amazon Bedrock
GCP	Vertex AI generative AI overview
Azure	Azure OpenAI Service overview
OpenAI	OpenAI API quickstart
Anthropic AI	Anthropic Claude documentation

Before using AI LLM features through Flashback, we recommend:

• Having at least one active AI provider account (e.g., OpenAI, Anthropic, AWS, GCP, or Azure). • Being able to create and manage API keys, secrets, and usage quotas. • Basic understanding of JSON, HTTP APIs, and prompt/response patterns. • Clarity on security practices (environment variables, key rotation, role-based access).

We recommend you to visit the documentation of your preferred AI providers to learn how to provision models, configure authentication, and monitor usage safely.

---

Future Platform Capabilities

The following sections describe the prerequisites that will help you prepare for upcoming Flashback features described in the “What’s Coming Next” roadmap.

AI Engine Aggregation

Prerequisites to get value from unified AI engine aggregation:

• Accounts and API keys for two or more AI providers (e.g., OpenAI, Anthropic, AWS Bedrock, Vertex AI, Azure OpenAI). • A clear mapping between projects/teams and the AI providers they are allowed to use (for policy-based routing). • Basic familiarity with model families (chat, embeddings, vision, tools/function calling) and their typical use cases. • Centralized way to store secrets (e.g., secret manager, vault, or environment management in your CI/CD pipeline).

Agentic Self-Management

Prerequisites to let agentic systems observe and optimize your workloads:

• Existing monitoring or logging setup for your applications (metrics, logs, traces), even if basic. • Clear SLOs or target KPIs for cost, latency, and reliability (e.g., max €/month, P95 latency, uptime %). • Permissions for Flashback to read observability data and, in controlled scopes, adjust policies or routing. • A staging/sandbox environment to safely test agent-driven changes before applying them to production.

Multi-Cloud Compute & Serverless Functions

Prerequisites to extend from storage into cross-cloud compute:

• Cloud accounts with permissions to create and manage compute resources (VMs, containers, serverless functions) on at least one provider. • A container registry or build pipeline if you deploy containerized workloads (e.g., Docker images). • Basic understanding of networking concepts (VPC/VNet, subnets, security groups, egress rules) on your main cloud. • SSH or API-based access patterns already defined for your applications, so Flashback can orchestrate them consistently.

Structured and Specialized Data Layers

Prerequisites to connect databases, warehouses, and analytics engines:

• At least one structured data system in use (e.g., PostgreSQL, MySQL, BigQuery, Snowflake, data warehouse, or OLAP engine). • Ability to create read-only and read-write roles on your databases or warehouses. • Network access patterns defined for data systems (e.g., private endpoints, IP allowlists, VPN/peering). • Basic understanding of SQL and/or analytical query patterns to define which workloads should be federated through Flashback.

DePIN-Native Governance & Incentives

Prerequisites to participate in decentralized governance and incentive mechanisms:

• A secure, organization-managed wallet solution (custodial or non-custodial) to interact with supported blockchains. • Internal policies for who can sign on-chain transactions and how keys are backed up and rotated. • Basic familiarity with staking, delegation, and reward distributions, especially for infrastructure-related tokens. • Compliance review for your jurisdiction regarding the use of digital assets in operations and incentives.