Master Generative AI with Java and Spring Boot | Spring AI

Explain large language models as the brain of AI, powered by neural networks that learn from data, recognize patterns, and predict text; show how developers use existing models in apps.

Learn how Spring AI enables easy AI integration in Spring Boot apps, with Spring AI documentation, chat and embedding APIs, plus vector databases, prompts, structured outputs, and function calls.

Explore how the OpenAI ChatGPT API calls the GPT-4 model via a chat interface, sending prompts and receiving responses. Implement this flow in a Java app using Spring AI.

Create a Spring Boot app that calls the OpenAI API with Spring AI. Set up via Spring Initializr, add Spring Web and OpenAI dependencies, and configure the API key.

Learn how to obtain and secure an OpenAI API key, configure billing and permissions, and integrate the OpenAI chat API with a Spring Boot application.

Configure the spring app to use the OpenAI chat API by setting the API key, organization id, and project id as environment variables and selecting the GPT-4 model for chat.

Create a Spring Boot rest controller exposing an API that takes a prompt and calls OpenAI's GPT-4 model via Spring AI chat client to return a response.

Call the chat client with spring eye functionality to access responses. Explore metadata, results, and the chat completion object, and use the prompt API for varied outputs.

Create an api that accepts a celebrity name and returns achievements by querying a ChatGPT model, using a prompt template with dynamic name replacement in Spring Boot with Spring AI.

Learn to externalize large prompts with spring templates in a java spring boot API, injecting classpath resources and reusing templates to generate prompts efficiently.

Learn how to structure prompts in a spring template API by defining system and user roles, passing prompts as messages, and using rules to guide a smart virtual assistant.

Develop an api in Spring Boot to retrieve a player's career achievements using a prompt-driven chat client, with a prompt template and response parsed into objects.

Master the use of Spring output converters to transform API data into bean, list, map, or bin formats, guided by prompts that define the required structure.

Create an API endpoint that returns a list of achievements for a given player name using an achievement entity and a list output converter.

Explore what prompt engineering is, learn how to craft effective prompts to maximize model output, discover prompt types, best practices, and practical demonstrations.

Discover prompt engineering by learning how to give clear instructions to an LM model to elicit efficient, varied outputs across formats. Explore how different prompt types shape results.

Explore the importance of prompt engineering for large language models, learn how to craft inputs that yield consistent outputs across models, and apply different prompt types to build reliable applications.

Use instruction-based prompts to give the model a direct task and produce clear outputs, from reversing a string to explaining why Rome's top attractions are popular.

Master completion prompts by leaving parts of a prompt open for the model to finish. Learn how fill-in-the-blank prompts guide the LLM to produce the intended output.

Explore zero-shot prompts as clear-cut, instruction-like queries without examples, and see how the model answers directly from the question, covering factorial, Newton's laws, and encapsulation in object oriented programming.

Few-shot prompts guide the model by providing a few examples to determine the next output, illustrated with arithmetic and loop coding examples.

Explains role-based prompts in prompt engineering and shows how assigning a specific role, such as chartered accountant or travel agent, guides model responses for targeted interview prep.

Explore conversational prompts that drive ongoing AI model interactions, using follow up questions and remembered context to build chat history and contextualized polymorphism-inspired answers.

Explore chain of thought prompts to reveal step-by-step reasoning behind LLM outputs, with math and coding examples like the fibonacci sequence and python explanations.

Leverage creative prompts to drive generative AI outputs across image, video, and audio tools, turning imaginative inputs into stories, math narratives, and travel scene descriptions.

learn how comparison prompts drive side-by-side analysis by asking the model to compare two or more items, revealing parameters like performance, ease of use, scalability, and security with table outputs.

Master generative ai with java and spring boot covers best practices in prompt engineering. It emphasizes clear, specific prompts, iterative refinements, retrieval augmented context, few-shot examples, and defining output formats.

Explore the challenges of prompt engineering and language models, including ambiguity, hallucination, and bias, and review applications in coding assistants, content creation, personalized learning, and automation.

Explore how Spring AI enables multimodal capabilities with OpenAI, processing image and audio: extract text from images, generate images from prompts, and transcribe audio or convert to speech in languages.

Build an image to text API using spring multimodal by creating an image controller and a chat model client to describe a given jpeg image, with prompts and user specs.

Describe prompts to generate images with the image generation API, configuring the model, prompt, image count, size, quality, and response format to return an image url.

Implement an audio to text workflow by integrating OpenAI's transcription API in a Spring Boot app, selecting models, and formatting outputs as text, json, or srt in multiple languages.

Explore building a text-to-speech endpoint in spring boot that uses OpenAI's tts model to convert prompts into mp3 audio, with configurable voice and speed, returned as a downloadable byte stream.

Learn how retrieval augmented generation uses external data and a relevancy system to keep generative AI outputs current, accurate, and well-sourced.

Explore how vector databases store non-structured data as vector embeddings and enable retrieval augmented generation in Spring Boot applications, using LM models for similarity search.

Implement rag in a Spring Boot app to build an API-based Indian Constitution simplification system, using vector embeddings, a vector store, and OpenAI LLMs for accurate, context-aware answers.

Build a Spring Boot app that converts a PDF of the Indian Constitution into a simple vector store using OpenAI embeddings, saving to a JSON file and reusing it.

Understand how tokens divide text into chunks for LLM models and embeddings, and how token counts drive API pricing and prompt usage. Explore how the tokenizer tool reveals token counts.

Load and reuse a vector store with embeddings to power the app, showing how to create, persist, and load it as a knowledge base for the API model.

Implement an api that uses a vector store to power retrieval augmented generation, wiring a controller, chat client, and advisors to fetch context via embeddings.

Rag workflow with a PG vector database in a Spring Boot app using Docker Compose, including adding the pg vector dependency, Docker Compose setup, and a vector store schema.

Configure a PG vector loader to initialize the database and load PDF embeddings into a PG vector store for a Spring Boot app, enabling similarity search.

Implement a rag workflow using a pg vector store in spring boot, performing similarity search to fetch documents and generate prompts with context.

Build a sentiment analysis app with Spring Boot backend and React frontend, using OpenAI to score feedback from minus one to plus one and store results in H2 in memory.

Create a frontend for a sentiment analysis system by scaffolding folders, building a React app with width, and configuring Axios calls, Tailwind CSS, and backend APIs.

Create a functional feedback UI by configuring tailwind, updating app.jsx, and building a styled textarea with a submit button. Implement a feedback history in subsequent steps.

Create a feedback history table in a simple UI styled with Tailwind CSS, with headers for feedback, score, and sentiment, and prepare for a Spring Boot backend.

Set up a Spring Boot back end for a sentiment analyzer using Maven, Spring Web, H2, JPA, Lombok, and OpenAI Starter, with OpenAI keys configured via environment variables.

Define a feedback entity with id, content, sentiment score, created at, and sentiment type; expose APIs to fetch all feedback and save entries in a spring boot app with h2.

Define a save feedback post API in a Spring Boot app to persist feedback and return saved entity, then compute sentiment and score via a Spring AI service with OpenAI.

Implement a frontend UI to save and fetch feedback via Axios post to /api/feedback, manage form state, and resolve CORS to persist sentiment score and type in the database.

Implement a feedback history interface by fetching data with the effect hook, rendering it in a table, and updating the list when new feedback is submitted, with sentiment color-coded.

Enhance the application's user interface by implementing a sentiment-based color scheme for table descriptions, using a get sentiment color function with positive, negative, and neutral cases to apply css classes.