Developer SDK

Build on Your Organization's Memory

The bRRAIn Platform SDK is a typed Go client for extensions that read and write the vault, call approved models, use MCP tools and integrations, and stay inside your organization's policies. Every organization gets five free developer seats.

import platformsdk "github.com/Qosil/bRRAIn/pkg/platform-sdk"

What the Platform SDK Gives You

Governed Vault Access

Read, write, search and manage your organization's knowledge, from Markdown notes to PDFs and images. Incoming content passes the organization's ingestion rules and security gate.

Models, Tools and Integrations

Call the models your organization has approved, with bRRAIn's own Handler as the default. Use MCP tools, live data connectors and vendor connections without storing a single credential.

Policy and Audit Built In

Ask the policy engine before acting, record every action in a tamper-evident audit trail, and keep each extension's data isolated in its own namespace.

Quick Start

Set up, connect, write and search. Full 7-step guide →

// go.mod: repository access comes with a developer seat
require github.com/Qosil/bRRAIn/pkg/platform-sdk v0.0.0-00010101000000-000000000000

replace github.com/Qosil/bRRAIn/pkg/platform-sdk => ../bRRAIn/pkg/platform-sdk
import platformsdk "github.com/Qosil/bRRAIn/pkg/platform-sdk"

c, err := platformsdk.New(platformsdk.Options{
	BaseURL: os.Getenv("BRRAIN_INTERNAL_URL"),
	Token:   os.Getenv("BRRAIN_INTERNAL_TOKEN"),
})
if err != nil {
	log.Fatal(err)
}
err = c.Vault().Write(ctx, platformsdk.VaultWriteRequest{
	Path:    "extensions/hello-notes/notes/first.md",
	Content: []byte("---\ntitle: \"First note\"\ntags: []\n---\n\nHello, vault.\n"),
})
if err != nil {
	log.Fatal(err)
}

res, err := c.Vault().Search(ctx, platformsdk.VaultSearchRequest{
	Query: "Hello",
	Path:  "extensions/hello-notes",
})
if err != nil {
	log.Fatal(err)
}
for _, h := range res.Hits {
	fmt.Println(h.Path, h.Snippet)
}

Built for every industry

Extensions put your organization's memory to work in the workflows each industry already runs.

Accounting Firms

Persistent AI memory gives every team member instant access to complete client engagement history.

Read the use case →

Professional Services

Team-wide contextual AI accelerates proposal turnaround and ensures engagement quality at scale.

Read the use case →

Enterprise IT

Persistent incident context and ITSM integration cut mean time to resolution dramatically.

Read the use case →

Marketing Firms

Persistent AI memory maintains campaign history, client preferences, and cross-channel insights across every engagement.

Read the use case →

Legal Firms

AI that remembers every precedent, every case pattern, and every client preference across the entire firm.

Read the use case →

Logistics Firms

Persistent memory that compounds route intelligence, customer preferences, and operational patterns across every shipment.

Read the use case →

Customer Service Firms

AI that knows every customer interaction, every resolution pattern, and every product issue — instantly.

Read the use case →

Tech Support Firms

Persistent AI memory that retains every incident, every resolution, and automatically updates institutional knowledge.

Read the use case →

Mining & Geological

Persistent AI memory ingests sensor data, bore logs, and geological samples into one queryable graph. Query by depth, mineral composition, or geological era across every site the company has ever touched.

Read the use case →

Healthcare

Persistent memory unifies the patient graph across every encounter. Find similar cases instantly, audit every AI suggestion, and keep PHI inside your own vault.

Read the use case →

IoT & Teleoperation

Time-series ingestion plus POPE graph over devices, sites, operators, and anomalies. Teleoperators inherit the full fleet memory on day one, and AI agents spot drift before it becomes an incident.

Read the use case →

Manufacturing

bRRAIn ingests every production event into a graph with lineage and dependency edges. Trace a defect from finished SKU to raw-material lot in milliseconds, and compound quality knowledge across plants.

Read the use case →

Architecture

Your extension runs on your organization's bRRAIn pod and reaches its platform services through one typed client, under the organization's rules.

Your extension Your code Routes, UI, jobs Platform SDK Typed Go client Vault & ingestion Models & MCP tools Policy & audit bRRAIn platform Your organization's pod Governed vault LLM Registry + Handler Policy engine · audit trail

What the SDK handles for you

Capability Building it yourself With the Platform SDK Governed by
Knowledge files, including PDFs and imagesYour own storageBuilt-inPer-extension namespaces
Bringing in outside contentManual checksBuilt-inYour ingestion rules and security gate
Model callsVendor keys in your codeBuilt-inYour LLM Registry; the Handler by default
Permission checksHand-rolledBuilt-inYour policy engine
Audit trailManualBuilt-inTamper-evident audit chain
Vendor credentialsStored by your appResolved when neededYour organization's connections