SHREE AI OS · DEVELOPER QUICKSTART

Developer
Quickstart.

Get started with Shree AI OS v1.0.6 Developer Preview in under 5 minutes. In-process AI runtime for Java with deterministic behavior, privacy-first boundaries, and 10 verified SDK facades on Java 21 LTS.

v1.0.6 · DEVELOPER PREVIEW

STEP 01 · PREREQUISITES

Java 21 LTS & Maven

Shree AI OS is native to the modern Java virtual machine. Virtual threads, record patterns, and sealed interfaces are leveraged for high-throughput cognitive pipelines.

Java Baseline

Java 21 LTS (or 24)

Build Tool

Maven 3.9+ / Gradle 8+

Vector Store

pgvector / In-Memory

STEP 02 · MAVEN DEPENDENCY

Add Dependency Coordinates

Add the official v1.0.6 developer preview coordinates to your pom.xml:

pom.xml

<!-- Maven Dependency Coordinates -->
<dependency>
    <groupId>io.github.darshanrathod04</groupId>
    <artifactId>shree-ai-os</artifactId>
    <version>1.0.6-developer-preview</version>
</dependency>

STEP 03 · RUNTIME BOOTSTRAP

Bootstrap with ShreeAI.builder()

Configure runtime security, default providers, and execute your first request through the 11-stage pipeline:

Application.java

package com.example.demo;

import com.shreeai.os.platform.sdk.ShreeAI;
import com.shreeai.os.platform.sdk.SDKResponse;
import com.shreeai.os.platform.core.RuntimeConfiguration;

public class Application {
    public static void main(String[] args) {
        // 1. Configure runtime with fail-closed security and hybrid RAG
        RuntimeConfiguration config = RuntimeConfiguration.builder()
            .defaultProvider("gemini-3.6-flash")
            .failClosed(true)
            .build();

        // 2. Initialize fluent client on Java 21 LTS
        ShreeAI shree = ShreeAI.builder()
            .apiKey(System.getenv().getOrDefault("GEMINI_API_KEY", "local"))
            .configuration(config)
            .build();

        // 3. Execute grounded chat through the 11-stage cognitive pipeline
        SDKResponse response = shree.chat("Explain hybrid RRF vector retrieval");
        System.out.println("Answer: " + response.answer());
        System.out.println("Confidence: " + response.confidence());

        // 4. Clean shutdown of virtual threads and kernel services
        shree.close();
    }
}

HYBRID RAG (KnowledgeSDK)

Vector Ingest & RRF Search

Combined HNSW vector proximity with GIN full-text keyword ranking:

KnowledgeExample.java

// Ingest documentation into pgvector with deduplication
shree.knowledge().ingest("PostgreSQL 16 with pgvector HNSW provides sub-millisecond retrieval.");

// Execute Hybrid RRF search (HNSW KNN + GIN Full-Text Search)
var results = shree.knowledge().search("pgvector HNSW performance");
results.forEach(entry -> System.out.println("Matched chunk: " + entry.getContent()));

EPISODIC CONTEXT (MemorySDK)

Tenant-Isolated Context

Retain and semantically recall context across multi-turn sessions:

MemoryExample.java

// Store episodic context scoped strictly to current tenant
shree.memory().store("deployment-target", "Kubernetes 1.30");

// Recall semantically relevant episodic context
var memories = shree.memory().recall("target cluster infrastructure");
memories.forEach(m -> System.out.println("Recalled: " + m.getValue()));

SHREE AI OS

Explore the 10 SDKs.
Or inspect the architecture.

Dive into the full SDK facade documentation or understand the 5-layer platform design.

Explore 10 SDKs↗Platform Architecture

DeveloperQuickstart.