Why Bob Is Bringing AI Into Spring Boot

Bob’s building a customer support system that can:

• Summarize messages
• Auto-reply using a language model
• Classify sentiment
• Generate product descriptions

Instead of building models from scratch, Bob uses:

• OpenAI API for LLMs (GPT-like magic)
• TensorFlow Java for local ML inference

Architecture Overview

[Client] ⬇️ [Spring Boot API] ⬇️⬇️ [OpenAI API] ← LLM tasks (chat, summarization) [TensorFlow] ← In-house models (sentiment, image tags)

Bob’s API acts as a smart gateway, combining cloud-based AI with local model inference.

Integrating OpenAI API in Spring Boot

Step 1: Add OpenAI Client (via WebClient or SDK)

@Configuration public class OpenAIConfig { @Bean public WebClient openAIClient() { return WebClient.builder() .baseUrl(“https://api.openai.com/v1”) .defaultHeader(“Authorization”, “Bearer ” + System.getenv(“OPENAI_API_KEY”)) .build(); } }

Step 2: Call GPT from a Service

@Service public class GPTService { @Autowired private WebClient openAIClient; public Mono getChatResponse(String prompt) { return openAIClient.post() .uri(“/chat/completions”) .bodyValue(Map.of( “model”, “gpt-3.5-turbo”, “messages”, List.of(Map.of(“role”, “user”, “content”, prompt)) )) .retrieve() .bodyToMono(String.class); } }

Supports:

• Summarization
• Text generation
• Translation
• Code generation

Integrating TensorFlow Java for Local ML Inference

Add TensorFlow Dependency

org.tensorflow tensorflow-core-platform 0.4.0

Example: Sentiment Classification

public class SentimentService { private SavedModelBundle model; public SentimentService() { model = SavedModelBundle.load(“models/sentiment”, “serve”); } public String predict(String inputText) { try (Tensor input = Tensors.create(inputText)) { Tensor result = model.session().runner() .feed(“input_text”, input) .fetch(“output_label”) .run().get(0); return result.expect(String.class).data().getObject(); } } }

Use cases:

• Image tagging
• Sentiment analysis
• Audio classification
• Anomaly detection

Combining AI Sources (OpenAI + TF)

Bob builds an endpoint that:

• Uses GPT to summarize a chat
• Then uses TensorFlow to classify sentiment
• Finally stores it in a database

@PostMapping(“/analyze”) public ResponseEntity analyze(@RequestBody String message) { String summary = gptService.getChatResponse(“Summarize: ” + message).block(); String sentiment = sentimentService.predict(message); return ResponseEntity.ok(new ChatAnalysis(summary, sentiment)); }

Best Practices

Security Tip

Use server-side proxying for GPT, not direct calls from frontend. This protects:

• Your OpenAI key
• Your business logic

Use Case Ideas

Find us

linkedin Shant Khayalian
Facebook Balian’s
X-platform Balian’s
web Balian’s
Youtube Balian’s

#springboot #openaiapi #tensorflowjava #javaml #machinelearning #gptintegration #aiinjava #springai #backendai #microservices