v1.0 — Now with LLM inference pipeline

Event processing
that thinks
for itself.

NextEvent is an AI-native event-driven framework for .NET 6. Build event-sourced systems with a built-in LLM inference pipeline — route, enrich, and act on events with context-aware intelligence.

Quick Start View on GitHub
MIT License
.NET 6 Runtime
AIO Bundle
NextEvent — AI Pipeline
$ dotnet add package NextEvent.AI
Installing NextEvent.AI v1.0.0...
Added package NextEvent.AI → 1 package(s).

$ builder.Services.AddNextEventPipeline()
Pipeline registered: EventIngest →
LLMEnricher → ActionRouter

$ // Event enriched by GPT-4o-mini inference
// NextEvent inferred: escalation_required=true
// NextEvent routed to: SLAHandler

Event processed in 38ms (p99)
Core Capabilities

Everything you need to ship event-driven AI systems.

From stream ingestion to LLM-powered routing — NextEvent bundles the full pipeline so your team ships faster without stitching together disparate libraries.

Event Sourcing

Append-only event store with snapshot support, optimistic concurrency, and configurable serialization (JSON, MessagePack, ProtoBuf).
EventStore.Core

LLM Inference Pipeline

Built-in enricher that runs OpenAI / Azure OpenAI / local LLMs over your event payload before routing — classify, extract, and augment in one pass.
AI-native

Multi-Transport Support

Consistent producer/consumer API across RabbitMQ, Apache Kafka, and Azure Service Bus. Swap transports with a single config line.
Kafka · RabbitMQ · ASB

Middleware Orchestration

Chain validators, transformers, and handlers with typed middleware — no magic strings, full DI integration, and first-class testing support.
Typed pipeline

Outbox Pattern

Reliable publish guarantee via the transactional outbox pattern. No lost events, no duplicate publishes — fully deterministic at-least-once delivery.
Reliability

Observability Built-In

Structured logging, OpenTelemetry traces, and Prometheus metrics off the shelf. Zero-config health checks for Kubernetes liveness/readiness probes.
OpenTelemetry
Documentation

Start building in minutes.

Program.cs — minimal setup 
// Install: dotnet add package NextEvent.AI
var builder = WebApplication.CreateBuilder(args);

// Register the AI-native pipeline
builder.Services.AddNextEventPipeline(config =>
{
    config.UseOpenAI(apiKey: builder.Configuration["OpenAI:Key"]);
    config.Model = "gpt-4o-mini";
});

var app = builder.Build();

// Subscribe to an event stream
app.MapPost("/events", async (EventEnvelope evt, IPipeline pipeline) =>
{
    var result = await pipeline.ProcessAsync(evt);
    return Results.Ok(result.RoutingDecision);
});

app.Run();

Quick Start Guide

5-minute hands-on

Architecture Deep Dive

How the pipeline works

Samples Repository

10+ runnable examples

API Reference

Full XML doc coverage
Changelog & Blog

What's new in NextEvent.

Release notes, deep-dives, and patterns from the team building it.

Release Jun 2026

v1.0 — AI Pipeline, Outbox, and OpenTelemetry GA

The first stable release ships the LLM inference pipeline, transactional outbox, and first-class OpenTelemetry instrumentation. Here's what changed and why.
by
brook yu · 8 min read
Deep Dive May 2026

Routing Events with LLMs — Without Adding Latency

We ran 50k events through GPT-4o-mini and kept p99 at 38ms. This is the architecture that made it possible: async enrichment, connection pooling, and smart batching.
by
brook yu · 12 min read
Pattern Apr 2026

Event Sourcing on .NET 6 — Lessons After 2 Years

Snapshot strategies, aggregate version conflicts, and the outbox pattern in practice. What we got right, what we'd do differently, and the .NET 6 specifics that matter.
by
brook yu · 15 min read
Open Source

Built in the open. Contributions welcome.

NextEvent is MIT-licensed and community-driven. Here's how to get involved.

brookyu/NextEvent_DotNet6_AIO
C# — stars

AI-native event processing framework for .NET 6. Event sourcing, LLM inference pipeline, multi-transport support, and transactional outbox.

MIT .NET 6 AI-native
1

Fork and clone the repo

Create your own copy and get the source locally.
git clone https://github.com/YOUR_HANDLE/NextEvent_DotNet6_AIO
2

Run the test suite

Verify everything passes before making changes.
dotnet test
3

Open a PR with your change

Describe what changed and link any issues. The team reviews within 48h.
About the Project

Built by developers, for developers.

NextEvent started as an internal tool for handling high-volume event streams with AI-powered routing. After three production iterations, we extracted it into an MIT-licensed framework so other .NET teams don't have to rebuild this from scratch.

Runtime .NET 6 LTS — back-compat to .NET 7/8
License MIT — free for commercial use
Maintainer brook yu — active on GitHub
Status v1.0 GA — stable API, active development
Maintainer
BY

brook yu

Creator & maintainer — NextEvent
@brookyu GitHub