Unlock Autonomous AI Agents with Spring AI Alibaba: Scheduling, Human‑in‑the‑Loop, and Real‑World Use Cases

Why Autonomous Agents?

Beyond the common chat‑based agent model, autonomous agents can continuously run by listening to environmental signals (timers, events, context changes) and initiate human‑machine interactions for tasks that require manual confirmation, balancing autonomy with controllability.

Spring AI Alibaba (SAA) Support

SAA provides a framework to quickly build such agents, offering scheduling APIs and management components for backend execution.

Business Impact

Autonomous agents expand individual capabilities, enable a single person plus multiple agents to collaborate on engineering tasks, and open scenarios like automated periodic business, batch clearing, emergency response, human‑in‑the‑loop decision making, long‑cycle tasks, and memory‑managed periodic jobs.

Key Scenarios

Automated Periodic Business : Scheduled agents collect and analyze data, generate visual reports, and improve report quality using LLMs.

Batch Processing : Parallel agents handle large‑scale financial data, news, and social media for investment insights.

Emergency Response : Agents monitor IoT or security events and trigger alerts automatically.

Human Decision Support : Agents perform routine analysis and involve humans only for high‑risk cases.

Long‑Cycle Tasks : Agents pre‑process multi‑modal data in the background, delivering ready‑to‑use results.

Memory Management : Periodic agents retain execution history to detect trends and route decisions.

Building a Timed Agent

Agents are registered via ScheduledAgentManager and can be scheduled with schedule(ScheduleConfig config). The default implementation runs within a single JVM process.

Example: Daily Report Agent

@Bean
public CompiledGraph dailyReportAgent(ChatModel chatModel) throws GraphStateException {
    ChatClient chatClient = ChatClient.builder(chatModel)
        .defaultAdvisors(new SimpleLoggerAdvisor())
        .build();
    AsyncNodeAction dataLoaderNode = node_async(state -> {
        // Load business report metadata (orders, products, feedback)
    });
    LlmNode llmDataAnalysisNode = LlmNode.builder()
        .chatClient(chatClient)
        .paramsKey("data_summary")
        .outputKey("summary_message_to_sender")
        .userPromptTemplate(DAILY_REPORT)
        .build();
    StateGraph stateGraph = new StateGraph("OperationAnalysisAgent", () -> {
        Map<String, KeyStrategy> strategies = new HashMap<>();
        strategies.put("data_summary", new ReplaceStrategy());
        strategies.put("summary_message_to_sender", new ReplaceStrategy());
        strategies.put("message_sender_result", new ReplaceStrategy());
        strategies.put("access_token", new ReplaceStrategy());
        return strategies;
    })
        .addNode("data_loader", dataLoaderNode)
        .addNode("data_analysis", node_async(llmDataAnalysisNode))
        .addNode("message_sender", node_async(generateMessageSender()))
        .addEdge(START, "data_loader")
        .addEdge("data_loader", "data_analysis")
        .addEdge("data_analysis", "message_sender")
        .addEdge("message_sender", END);
    CompiledGraph compiledGraph = stateGraph.compile();
    compiledGraph.setMaxIterations(100);
    ScheduleConfig scheduleConfig = ScheduleConfig.builder()
        .cronExpression("0 0 8 */1 * ?") // every day at 8 AM
        .build();
    compiledGraph.schedule(scheduleConfig);
    return compiledGraph;
}

Example: Evaluation Analysis Agent

@Bean
public CompiledGraph evaluationAnalysisAgent(ChatModel chatModel, FeedbackMapper feedbackMapper) throws GraphStateException {
    ChatClient chatClient = ChatClient.builder(chatModel)
        .defaultAdvisors(new SimpleLoggerAdvisor())
        .build();
    EvaluationClassifierNode sessionAnalysis = EvaluationClassifierNode.builder()
        .chatClient(chatClient)
        .inputTextKey("iterator_item")
        .outputKey("session_analysis_result")
        .categories(List.of("yes", "no"))
        .classificationInstructions(List.of(
            "Return pure JSON with fields: user, time, complaint, satisfaction, summary.",
            "complaint: indicates if the feedback is a complaint (yes/no).",
            "satisfaction: actual user satisfaction.",
            "summary: core complaint point and improvement direction"
        ))
        .build();
    // Build state graph, nodes, edges, conditional logic, and schedule (hourly)
    // ... (omitted for brevity) ...
    ScheduleConfig scheduleConfig = ScheduleConfig.builder()
        .cronExpression("0 0 */1 * * ?") // every hour
        .build();
    compiledGraph.schedule(scheduleConfig);
    return compiledGraph;
}

Conclusion

Autonomous AI agents expand enterprise intelligence by combining scheduled triggers, event‑driven responses, and human‑in‑the‑loop mechanisms. Spring AI Alibaba offers a ready‑to‑use framework for developers to create custom agents that automate data collection, analysis, and decision‑making end‑to‑end.