// author

Navmeet Kaur

Navmeet writes about software architecture for the AI era — how agentic systems are actually designed, not just demoed. Her work covers AI-native application architecture, agent orchestration and memory, context engineering, and where AI agents fit alongside (and eventually replace) traditional services and microservices. She focuses on the design decisions that survive production — state, tools, boundaries, and failure modes — turning fast-moving AI patterns into architecture developers can build on with confidence.

23 guides~2.1 hrs totalupdated Jul 6, 2026

 Start here

Agent Mesh vs Supervisor: What Holds Up in Production

Agent mesh vs supervisor: why 2026 production data favors bounded supervisor coordination over open agent meshes, and when a controlled mesh still earns its keep.

Intermediate · 6 min

Software Architecture

Read the guide →

All guides

23 on this page

BeginnerCDNs and Object Storage: Store at the Origin, Serve at the Edge (2026)CDNs and object storage explained in 2026 — object storage as the origin for large files, CDNs as the edge cache near users, how they work together, and what they store in AI systems.5 min BeginnerRate Limiting Explained: Token Bucket, Leaky Bucket, and When Each Wins (2026)Rate limiting explained in 2026 — fixed window, sliding window, token bucket, and leaky bucket, which algorithm to default to, and why LLM traffic is limited by tokens rather than requests.5 min BeginnerHorizontal vs Vertical Scaling: Scale Up Until You Can't (2026)Horizontal vs vertical scaling in 2026 — scale up (a bigger box) versus scale out (more boxes), the hardware ceiling, why horizontal needs statelessness, and the diagonal middle ground.5 min BeginnerCAP Theorem Explained: It Was Never 'Pick Two' (2026)CAP theorem explained in 2026 — why 'pick two of three' is wrong, the real consistency-vs-availability choice during a partition, PACELC, and what it means for AI systems.5 min BeginnerEvent-Driven Architecture: Trading Control for Decoupling (2026)Event-driven architecture explained in 2026 — events vs commands, choreography vs orchestration, and the real trade-off: you gain decoupling and lose the single view of what happens next.5 min BeginnerAPI Gateway Explained: It's Not a Load Balancer (2026)API gateway explained in 2026 — what it actually does (auth, rate limiting, routing, aggregation), why it's not a load balancer, and where the new AI gateway fits.4 min BeginnerSQL vs NoSQL: It's About Access Patterns, Not Scale (2026)SQL vs NoSQL in 2026 — why the real choice is about access patterns and correctness, not scale; the four NoSQL types; and why most systems end up using both.5 min BeginnerLoad Balancers Explained: Health Checks Do the Real Work (2026)Load balancers explained in 2026 — L4 vs L7, the algorithms, and why health checks (not even distribution) are the part that actually keeps you online.5 min BeginnerSystem Design Building Blocks: A Decision Map, Not a Parts List (2026)System design building blocks explained as a decision map — the five questions that pick your components (architecture style, cache, message broker, database, load balancer) and how each block earns its place.6 min BeginnerCaching Strategies: Why Invalidation Is the Hard Part (2026)Caching strategies in 2026 — the patterns (cache-aside, write-through, write-behind), eviction policies, and why cache invalidation, not adding a cache, is the genuinely hard part.7 min BeginnerKafka vs RabbitMQ: A Log vs a Queue, Not Fast vs Slow (2026)Kafka vs RabbitMQ in 2026 — why it's a log versus a queue (not fast versus slow), the differences that matter, and when each one (or both) is the right call.7 min BeginnerMonolith vs Microservices: It's About Team Size, Not Traffic (2026)Monolith vs microservices in 2026 — the differences that matter, why team size (not traffic) decides it, and when to start with a modular monolith.7 min AI Gateway Design: One Front Door for Every Model (2026)AI gateway design: one front door for every LLM call — routing, fallback, caching, cost control, and guardrails. What belongs in one, and when you need it.8 min BeginnerAI Engineer Salary 2026: Real Numbers by LevelWhat's a real AI engineer salary in 2026? Numbers by level, the true premium over software engineers, and why one big figure can mislead you.5 min BeginnerWill AI Replace Software Engineers? What the 2026 Data SaysWill AI replace software engineers? The 2026 data says no in aggregate, but yes for the junior rung. What's really changing, and what to do.5 min IntermediateAgent Orchestration Patterns: A 2026 GuideAgent orchestration patterns: supervisor, hierarchical, sequential, parallel, swarm — which wins in 2026, what it costs, and when one agent is enough.5 min IntermediateAI Control Plane Architecture: A 2026 GuideAI control plane architecture: the governance layer over your agents — identity, policy, observability, cost, and audit — and when you actually need one.5 min IntermediateHuman-in-the-Loop Architecture: A 2026 GuideHuman-in-the-loop architecture for AI agents: approval gates, interrupt and resume, calibrated autonomy, the approval-fatigue trap, and where to put the human.5 min IntermediateLong-Running AI Workflows: A 2026 GuideLong-running AI workflows: why request/response breaks, how durable execution resumes instead of restarting, idempotency, and when to skip it.5 min IntermediateAI Agent Memory Architecture: A 2026 GuideAI agent memory for architects: the three long-term types, the write-recall-forget lifecycle, where it breaks, and when an agent needs no memory at all.6 min IntermediateAI Agents vs Microservices: An Architect's Guide (2026)AI agents vs microservices: the six differences that matter — determinism, state, contracts, retries, testing, cost — and when an agent is the wrong tool.6 min IntermediateContext Engineering Architecture: A 2026 GuideContext engineering for architects: treat the context window as RAM on a token budget — what fills it, where it breaks, and when a plain prompt wins.5 min

Following Navmeet Kaur?

Get new Navmeet Kaur guides the day they publish. One email, no spam.