LangChaingo API - Java SDK Documentation

BosBase exposes the /api/langchaingo endpoints so you can run LangChainGo powered workflows without leaving the platform. The Java SDK wraps these endpoints with the pb.langchaingo service.

The service exposes four high-level methods:

Method	HTTP Endpoint	Description
`pb.langchaingo.completions()`	`POST /api/langchaingo/completions`	Runs a chat/completion call using the configured LLM provider.
`pb.langchaingo.rag()`	`POST /api/langchaingo/rag`	Runs a retrieval-augmented generation pass over an `llmDocuments` collection.
`pb.langchaingo.queryDocuments()`	`POST /api/langchaingo/documents/query`	Asks an OpenAI-backed chain to answer questions over `llmDocuments` and optionally return matched sources.
`pb.langchaingo.sql()`	`POST /api/langchaingo/sql`	Lets OpenAI draft and execute SQL against your BosBase database, then returns the results.

Each method accepts an optional model configuration:

Map<String, Object> modelConfig = new HashMap<>();
modelConfig.put("provider", "openai");  // or "ollama" or other provider
modelConfig.put("model", "gpt-4o-mini");
modelConfig.put("apiKey", "your-api-key");  // Optional, overrides server defaults
modelConfig.put("baseUrl", "https://api.openai.com/v1");  // Optional

If you omit the model section, BosBase defaults to provider: "openai" and model: "gpt-4o-mini" with credentials read from the server environment. Passing an apiKey lets you override server defaults on a per-request basis.

Text + Chat Completions

import com.bosbase.sdk.BosBase;
import java.util.*;

BosBase pb = new BosBase("http://localhost:8090");

Map<String, Object> modelConfig = new HashMap<>();
modelConfig.put("provider", "openai");
modelConfig.put("model", "gpt-4o-mini");

List<Map<String, Object>> messages = new ArrayList<>();
Map<String, Object> systemMsg = new HashMap<>();
systemMsg.put("role", "system");
systemMsg.put("content", "Answer in one sentence.");
messages.add(systemMsg);

Map<String, Object> userMsg = new HashMap<>();
userMsg.put("role", "user");
userMsg.put("content", "Explain Rayleigh scattering.");
messages.add(userMsg);

Map<String, Object> request = new HashMap<>();
request.put("model", modelConfig);
request.put("messages", messages);
request.put("temperature", 0.2);

ObjectNode completion = pb.langchaingo.completions(request);
System.out.println(completion.get("content").asText());

The completion response mirrors the LangChainGo ContentResponse shape, so you can inspect the functionCall, toolCalls, or generationInfo fields when you need more than plain text.

Retrieval-Augmented Generation (RAG)

Pair the LangChaingo endpoints with the /api/llm-documents store to build RAG workflows. The backend automatically uses the chromem-go collection configured for the target LLM collection.

Map<String, Object> ragRequest = new HashMap<>();
ragRequest.put("collection", "knowledge-base");
ragRequest.put("question", "Why is the sky blue?");
ragRequest.put("topK", 4);
ragRequest.put("returnSources", true);

Map<String, Object> filters = new HashMap<>();
Map<String, Object> where = new HashMap<>();
where.put("topic", "physics");
filters.put("where", where);
ragRequest.put("filters", filters);

ObjectNode answer = pb.langchaingo.rag(ragRequest);
System.out.println(answer.get("answer").asText());

ArrayNode sources = (ArrayNode) answer.get("sources");
if (sources != null) {
    for (JsonNode source : sources) {
        System.out.println("Score: " + source.get("score").asDouble());
        System.out.println("Title: " + source.path("metadata").path("title").asText());
    }
}

Set promptTemplate when you want to control how the retrieved context is stuffed into the answer prompt:

Map<String, Object> ragRequest = new HashMap<>();
ragRequest.put("collection", "knowledge-base");
ragRequest.put("question", "Summarize the explanation below in 2 sentences.");
ragRequest.put("promptTemplate", "Context:\\n{{.context}}\\n\\nQuestion: {{.question}}\\nSummary:");

ObjectNode answer = pb.langchaingo.rag(ragRequest);

LLM Document Queries

Note: This interface is only available to superusers.

When you want to pose a question to a specific llmDocuments collection and have LangChaingo+OpenAI synthesize an answer, use queryDocuments. It mirrors the RAG arguments but takes a query field:

// Authenticate as superuser first
pb.admins().authWithPassword("admin@example.com", "password", null, null, null, null, null, null, null, null);

Map<String, Object> queryRequest = new HashMap<>();
queryRequest.put("collection", "knowledge-base");
queryRequest.put("query", "List three bullet points about Rayleigh scattering.");
queryRequest.put("topK", 3);
queryRequest.put("returnSources", true);

ObjectNode response = pb.langchaingo.queryDocuments(queryRequest);
System.out.println(response.get("answer").asText());
System.out.println(response.get("sources"));

SQL Generation + Execution

Important Notes:

This interface is only available to superusers. Requests authenticated with regular users tokens return a 401 Unauthorized.

It is recommended to execute query statements (SELECT) only.

Do not use this interface for adding or modifying table structures. Collection interfaces should be used instead for managing database schema.

Directly using this interface for initializing table structures and adding or modifying database tables will cause errors that prevent the automatic generation of APIs.

Superuser tokens (_superusers records) can ask LangChaingo to have OpenAI propose a SQL statement, execute it, and return both the generated SQL and execution output.

// Authenticate as superuser
pb.admins().authWithPassword("admin@example.com", "password", null, null, null, null, null, null, null, null);

Map<String, Object> sqlRequest = new HashMap<>();
sqlRequest.put("query", "Add a demo project row if it doesn't exist, then list the 5 most recent projects.");
sqlRequest.put("tables", List.of("projects"));  // optional hint to limit which tables the model sees
sqlRequest.put("topK", 5);

ObjectNode result = pb.langchaingo.sql(sqlRequest);
System.out.println(result.get("sql").asText());    // Generated SQL
System.out.println(result.get("answer").asText()); // Model's summary of the execution
System.out.println(result.get("columns"));
System.out.println(result.get("rows"));

Use tables to restrict which table definitions and sample rows are passed to the model, and topK to control how many rows the model should target when building queries. You can also pass the optional model block described above to override the default OpenAI model or key for this call.

Complete Example

import com.bosbase.sdk.BosBase;
import java.util.*;

public class LangChaingoExample {
    public static void main(String[] args) throws Exception {
        BosBase pb = new BosBase("http://localhost:8090");
        
        // Example 1: Chat completion
        Map<String, Object> modelConfig = new HashMap<>();
        modelConfig.put("provider", "openai");
        modelConfig.put("model", "gpt-4o-mini");
        
        List<Map<String, Object>> messages = new ArrayList<>();
        Map<String, Object> userMsg = new HashMap<>();
        userMsg.put("role", "user");
        userMsg.put("content", "What is the capital of France?");
        messages.add(userMsg);
        
        Map<String, Object> completionRequest = new HashMap<>();
        completionRequest.put("model", modelConfig);
        completionRequest.put("messages", messages);
        
        ObjectNode completion = pb.langchaingo.completions(completionRequest);
        System.out.println("Completion: " + completion.get("content").asText());
        
        // Example 2: RAG query
        Map<String, Object> ragRequest = new HashMap<>();
        ragRequest.put("collection", "knowledge-base");
        ragRequest.put("question", "Why is the sky blue?");
        ragRequest.put("topK", 3);
        ragRequest.put("returnSources", true);
        
        ObjectNode ragAnswer = pb.langchaingo.rag(ragRequest);
        System.out.println("RAG Answer: " + ragAnswer.get("answer").asText());
    }
}