Engineering real-time.
Architecture deep dives, tutorials, and lessons learned building real-time infrastructure at scale.
Connect Your Local LLM to NoLag in 5 Minutes
Wire up Ollama as a NoLag worker agent in under 5 minutes. No framework, no boilerplate -- just a transport layer between your model and the rest of your system.
Hybrid LLM Coordination: Mixing Local and Proprietary Models
Route tasks to the right model: local LLMs for speed and privacy, proprietary APIs for reasoning. NoLag coordinates the handoffs without coupling your agents to any single provider.
Why Multi-Agent Systems Need a Coordination Layer
Most teams building multi-agent systems reinvent coordination plumbing on Redis or Kafka. Here is why pub/sub is the right primitive and how six patterns cover the coordination needs.
Building an Agentic Chatbot with NoLag
Step-by-step tutorial: combine @nolag/chat for the user-facing interface with @nolag/agents for AI-powered responses. Full code walkthrough with escalation and monitoring.
From Human Realtime to Agent Realtime
How NoLag's real-time messaging infrastructure maps to agent coordination. Topics become task queues, rooms become workflow boundaries, and presence becomes health monitoring.
WebSockets vs Polling: Why Real-Time Apps Need a Persistent Connection
HTTP polling wastes bandwidth and adds latency. WebSockets hold the line open. Here is when each approach makes sense, and why most teams eventually move to WebSockets.
How to Build a Real-Time Chat App in Under 10 Minutes
A step by step walkthrough of building multi-room chat with typing indicators, presence, and message history using a purpose-built chat SDK.
MessagePack vs JSON: 47% Smaller Payloads for Real-Time Apps
JSON is human readable. MessagePack is wire efficient. We benchmarked both across chat, IoT, and dashboard workloads to see how much bandwidth you actually save.
The Pub/Sub Pattern Explained: Topics, Rooms, and Message Routing
Publish/subscribe is the backbone of every real-time system. This guide covers how topics, rooms, and filters work together to route messages at scale.
QoS 0, 1, and 2: Choosing the Right Delivery Guarantee for Every Message
Not every message needs the same delivery guarantee. Telemetry can afford to drop a packet. Payments cannot. Here is how per-message QoS works in practice.
Building Presence at Scale: Who Is Online and How to Track It
Presence looks simple until you have 10,000 concurrent users across hundreds of rooms. This post covers the architecture behind scalable online/offline tracking.
Real-Time Dashboards Without the Firehose: Filtered Subscriptions
Most dashboards show a subset of data, but most pub/sub systems deliver everything. Topic filters solve this by subscribing only to the entities on screen.
IoT Fleet Tracking Over WebSockets: Architecture for 100K Devices
GPS trackers, sensors, and connected vehicles generate a constant stream of telemetry. Here is how to architect a fleet tracking system that scales.
Handling Disconnects Gracefully: Reconnection and Message Replay
Mobile networks drop. Wi-Fi switches. Laptops close. A good real-time system recovers silently. This post covers reconnection strategies and message replay.
Blueprint SDKs vs Raw Pub/Sub: Ship in Minutes Instead of Weeks
Every real-time app rebuilds the same plumbing: rooms, presence, typing, replay. Blueprint SDKs give you a domain specific API so you can skip the wiring.