Building AI Chatbots in ASP.NET Core

using System.Text;
using Newtonsoft.Json;

public class GPTService
{
    private readonly HttpClient _httpClient;
    private readonly IConfiguration _configuration;

    public GPTService(HttpClient httpClient, IConfiguration configuration)
    {
        _httpClient = httpClient;
        _configuration = configuration;
    }

    public async Task<string> GetResponseAsync(string prompt)
    {
        var apiKey = _configuration["OpenAI:ApiKey"];
        var model = _configuration["OpenAI:Model"];
        var endpoint = _configuration["OpenAI:BaseUrl"];

        _httpClient.DefaultRequestHeaders.Clear();
        _httpClient.DefaultRequestHeaders.Add("Authorization", $"Bearer {apiKey}");

        var requestBody = new
        {
            model,
            messages = new[]
            {
                new { role = "user", content = prompt }
            }
        };

        var json = JsonConvert.SerializeObject(requestBody);
        var content = new StringContent(json, Encoding.UTF8, "application/json");

        var response = await _httpClient.PostAsync(endpoint, content);
        response.EnsureSuccessStatusCode();

        return await response.Content.ReadAsStringAsync();
    }
}

Step 6: Register Services

Update Program.cs:

builder.Services.AddHttpClient();
builder.Services.AddScoped<GPTService>();

Step 7: Create the API Controller

Create Controllers/GPTController.cs:

using Microsoft.AspNetCore.Mvc;

[ApiController]
[Route("api/[controller]")]
public class GPTController : ControllerBase
{
    private readonly GPTService _gptService;

    public GPTController(GPTService gptService)
    {
        _gptService = gptService;
    }

    [HttpPost]
    public async Task<IActionResult> Ask(GPTRequest request)
    {
        var result = await _gptService.GetResponseAsync(request.Prompt);
        return Ok(result);
    }
}

Step 8: Test the Endpoint

Send a POST request.

Endpoint

/api/gpt

Request Body

{
  "prompt": "Explain dependency injection in .NET"
}

Response

{
  "content": "Dependency Injection is a design pattern..."
}

Step 9: Production Best Practices

To build a secure and scalable solution:

• Secure API keys using environment variables or secret managers.

• Rate limit requests to prevent misuse.

• Log and monitor with Serilog or Application Insights.

• Cache responses for repeated queries to save costs.

Real-World Use Cases

OpenAI GPT can power a wide range of applications:

• Chatbots – customer support and virtual assistants

• Copilots – developer helpers and productivity tools

• Document intelligence – summarization and analysis

• Knowledge base search – semantic search with vector databases

• Content generation – technical documentation, reports, and blogs

Summary

OpenAI GPT can be integrated into ASP.NET Core applications through a simple service layer that uses HttpClient to communicate with the API. By configuring API settings, defining request and response models, registering services through dependency injection, and exposing REST endpoints, developers can quickly add conversational AI capabilities to their applications. With proper security, monitoring, caching, and rate limiting, this integration can support production-grade chatbots, copilots, document intelligence solutions, and enterprise AI assistants.