From theory to the systems that demand real understanding.
Long-form pieces on how production-AI teams orchestrate agents, serve models, and build the stack, plus GATE DA essays that turn probability, linear algebra, DBMS, ML, and AI into durable concepts.
When not to use RAG
RAG is the default answer to 'how do I give my LLM custom knowledge?' — and for a wide class of problems, it's the wrong answer. Long-context prompt caching, fine-tuning, and in-context learning each beat RAG in regimes where the costs and assumptions actually pencil out. Here's the decision tree.
The workplace health gap: menstruation, fertility, and menopause
Menstruation, fertility, and menopause quietly shape the careers of much of the workforce, and the silence around them pushes experienced women out the door.
Claude Code, Cursor, and Aider — three answers to the same question
Three coding agents, three different bets on autonomy, context, and where the human belongs in the loop. The interesting part isn't which one wins — it's which assumptions each one refused to compromise on.
Replit Agent's architecture, two years in
Replit Agent launched in September 2024 and turned a nine-year-old IDE into a $150M-ARR business. The architecture is unfashionably explicit — plan, confirm, execute, checkpoint — and the bet that explicit checkpoints beat full autonomy is paying off. Here's how it actually works.
MCP is the USB-C of AI tools, finally
Anthropic shipped the Model Context Protocol in late 2024 as a small, JSON-RPC-shaped spec for connecting LLM clients to tools and data sources. Eighteen months later, it's the most adopted open protocol in the AI space — and the hype is, for once, an undercount.
v0 by Vercel: how prompt-to-UI actually works
v0 makes 'build me a dashboard' produce real Next.js + shadcn/ui code that ships. The trick wasn't a smarter model — it was a narrow constraint, a streaming preview loop, and an opinionated component library acting as scaffolding. Here's the architecture, the competitive landscape, and why most clones missed the point.
AI SREs in production: Resolve.ai, Cleric, Parity
A new category of on-call agent is being shipped into real production environments — agents that watch alerts, gather evidence, and propose remediations. Most of them are deliberately read-only. The architecture is converging, the MTTR numbers are real, and PagerDuty is responding from the incumbent side. Here's what's actually working.
Speculative decoding in the wild: how labs cut latency by 2-3x
A small fast model proposes, the big slow model verifies. Across Medusa, EAGLE, and draft-model approaches, speculative decoding is the latency-reduction lever that frontier labs reach for when they can't make the model itself any smaller.
Why multi-agent swarms keep failing to ship
Two years after AutoGen and CrewAI promised teams of cooperating AI agents, the production scoreboard is brutal: orchestrator-workers ships, agent teams mostly don't. The reasons are structural, not stylistic — and they explain why every serious agent team has quietly converged on a much more boring shape.
Getting tool calling to 99% reliability in production
Most production agent failures are not model failures. They are tool-call failures — the model picked the right tool but wrote the wrong arguments, or hallucinated an argument that doesn't exist. The fix is unglamorous and mostly about schema design.
Voice agents at scale: the Vapi, Retell, Bland.ai engineering
Voice agents are the unsexy success story of the agent era. Underneath the marketing they're all the same five-box pipeline — STT, LLM, TTS, turn-taker, telephony — fighting for the same 500ms latency budget. Here's how the three biggest platforms actually build it, what each one optimizes for, and where the real cost goes.
Devin's architecture, anatomised
Cognition's launch demo promised a 'fully autonomous software engineer.' The product that actually ships is more interesting — a planner-executor split running in a sandboxed VM, with explicit memory windows, hard pruning rules, and a shockingly disciplined view of where the model is allowed to be creative.
We built the same agent three times: MAF, LangGraph, and ADK
Three frameworks, one problem: an invoice-processing agent that ingests PDFs, validates against a database, asks a human when uncertain, and writes back to an ERP. What each framework made easy. What each one made painful.
Sierra's customer-service playbook
Bret Taylor and Clay Bavor built the highest-profile agent company by making three opinionated bets — every brand needs its own AI, agents must run inside the brand's workflows, and pricing should track outcomes. Two years on, the bets are paying off in ways the industry is still copying.
AutoGPT to 2026: what survived
Three years after AutoGPT briefly broke GitHub's star counter, almost nothing in its original form has shipped. The interesting question is what did survive — and it's a sharper, more honest answer than the discourse of 2023 ever got close to.
Self-RAG and Corrective RAG, the loops that actually help
Two papers from 2023 argued RAG should be reflective: retrieve, judge, re-query if the evidence is bad. By 2026 the practitioner verdict is in — Self-RAG and CRAG actually do help, but only on the queries that need them, and the latency cost is real.
Bedrock, Vertex, Foundry: pick the one whose spirit matches yours
Every comparison of the three hyperscaler AI platforms is a feature matrix that nobody reads twice. The actual decision is about which company's spirit your team can live with for the next five years — and that's a different question.
Continuous batching: the one trick that made LLM serving 10x cheaper
Static batching wastes GPUs by padding to the longest sequence and waiting for stragglers. Continuous batching reschedules every iteration, evicts finished requests, and admits new ones mid-stream. Here's why that single shift bought a 10-23x throughput win.
How Cursor's Composer actually works
Multi-file edits feel atomic in Cursor not because the model got smarter, but because the team built a stack of careful workarounds — speculative diffs, a separate Apply Model, and an indexer that stays one step ahead of you. Here's the engineering.
Computer-use latency engineering: getting browser agents under a second
A naive browser-use loop is 4-8 seconds per step. Production systems run at 600ms. The gap is closed by half a dozen techniques — differential screenshots, prompt caching, batched actions, vision-model routing — each of which sounds boring until you measure the difference.
Google's Agent Development Kit, and when to reach for it
ADK is Google's bet on agent infrastructure: a thin SDK that takes Gemini's strengths — tool calling, grounding, multimodality — and makes them deployable to Vertex AI Agent Engine in one command. It is less framework than LangGraph, more framework than calling the model directly, and uniquely useful inside Google Cloud.
Late interaction, or why ColBERT keeps coming back
Single-vector dense retrieval crushes one nuance per chunk into one point in space. Late-interaction models like ColBERT keep a vector per token and decide similarity at query time. The math is beautiful. The storage cost is brutal. Here's when it's worth it.
Sub-agents, handoffs, supervisors — pick exactly one
Three topologies dominate multi-agent systems in 2026, and teams keep mixing them in ways that break debuggability. Sub-agents are parallel scoped work that returns to a parent. Handoffs are baton-passing with no return. Supervisors are explicit central routers. They are not interchangeable. Here's the framework for picking, and the failure modes when you don't.
Prompt caching: the 90% cost cut explained
Anthropic, OpenAI, DeepSeek and Google all ship prompt caching now, but the pricing models diverge sharply. For a tool-using agent with a long system prompt, getting the cache pattern right is the difference between a viable product and a P&L disaster.