Your AI Bill Just Spiked: How to Debug It and Pinpoint the Cause
The invoice email lands and your AI spend is 4× last week. The instinct is to…
Tag
#llm
44 posts
AI Application Performance: Stream First, Shrink the Context, Route the Model
LLM applications have two performance budgets that classic backends don’t worry…
The AI Engineering Vocabulary: Evals, Guardrails, RAG and the Words That Actually Matter
Every field has a vocabulary you have to own before anything else makes sense…
Building an AI FAANG Interview Coach: Run the Code, Don't Let the LLM Grade It
Ask ChatGPT to “review this solution and tell me if I’d pass the interview” and…
Building an AI Astrology (Kundali) Chatbot: Let the LLM Talk, Let the Ephemeris Do the Math
Ask ChatGPT “where was Saturn when I was born, 2:30 pm on 15 August 1990 in New…
The LLM Brain for Voice
Between the ears (STT) and the voice (TTS) sits the brain: the LLM that decides…
Anatomy of a Voicebot
Before any transports, providers or frameworks, you need one clear picture in…
Open LLMs in Production — Serving, Quantization, Fine-Tuning and Cost
Open-weight language models have become genuinely good. Llama, Qwen, Mistral…
The Real Cost in Rupees
A founder once told me, proudly, that self-hosting had cut their LLM costs by 8…
Evaluation and Safe Rollout
A team swapped their self-hosted model from FP16 to a 4-bit quantized version to…
Scaling on Kubernetes
A team moved their working single-GPU LLM service onto Kubernetes for…
Quantization: Smaller Weights, Measured Error
“We quantized the model to 4-bit and it got dumb.” I’ve heard this many times…
Fine-Tuning with LoRA and QLoRA
A team spent six weeks fine-tuning a model to answer questions about their…
Memory Maths: Weights and KV Cache
A team once told me their 8B model “needs 80 GB of GPU memory” because it kept…
Latency: TTFT, TPOT and Prefix Caching
A product team reduced their chatbot’s total response time by 40% and users…
Serving Engines: vLLM, llama.cpp and PagedAttention
A team I knew wrote their first LLM service as a FastAPI endpoint that called…
Should You Self-Host an Open LLM?
A fintech team I advised was about to sign a large annual contract with an LLM…
How LLM Inference Uses Your Hardware
When people first benchmark an LLM, they’re often confused by a result like this…
Batching and Throughput
Here’s a fact that confuses people at first: a GPU serving one user and a GPU…
Security, Compliance + Capstone
We’ve covered how to choose, size, speed up, cost out, fine-tune, evaluate and…
AI Agent Development — From First Loop to Production
Every second LinkedIn post today is about AI agents. Most of them show a demo…
Securing AI Agents — Prompt Injection, Least Privilege, and Sandboxing
In security reviews for agents, I now start with one question: “If this agent…
Debugging and Observability for AI Agents
A support executive pinged me on a Saturday morning: “The bot told a customer…
Memory and Context Management for Agents
When the first million-token context windows arrived, a friend of mine (very…
Cost Engineering for AI Agents
The first time an agent bill really scared me was during a festive sale. We had…
What an AI Agent Really Is (and When Not to Build One)
A few months back, a product manager from a lending startup in Gurugram called…
AI Agents Demystified — It's Just Automation With a Better Brain
Let’s cut through the noise. If you read Twitter or LinkedIn, you’d think “AI…
LLM Engineering in Production
Most LLM tutorials stop at “call the API and print the result.” Production doesn…
Prompt Injection — Attacks and Defenses
Prompt injection is the most important security vulnerability in LLM…
Building AI Agents with Tool Use
An LLM by itself can only generate text. It cannot check the weather, query a…
Securing Your LLM Application
Traditional web applications have well-understood security models: validate…
Streaming Responses in Production
A user types a question and clicks send. Without streaming, they stare at a…
Cost Optimization for LLM Applications
Your LLM prototype costs $2/day. Your production system serving 10,000 users…
Evaluating LLM Output Quality
Here is a scenario every LLM engineer has lived through. You tweak a prompt to…
Choosing Between OpenAI, Claude, and Open Source
Choosing a model is one of the first decisions you make when building an LLM…
LLM Observability and Monitoring
Traditional application monitoring asks: “Is the server up? Is latency…
Advanced Prompting — Chain of Thought, Few-Shot
In Lesson 3 you learned the fundamentals — system prompts, output formatting…
How LLMs Actually Work — What You Need to Know
You are going to build production systems on top of LLMs. Before you do, you…
Prompt Engineering Fundamentals
Prompt engineering is your primary interface to the model. In a traditional…
Choosing Models & Building Production AI Apps
You have built prototypes. You have called APIs, written prompts, wired up RAG…
LLMs for Programmers — How AI APIs Actually Work
Before you write a single line of AI code, you need a mental model of what you…
Calling LLM APIs — OpenAI, Anthropic & Structured Outputs
This lesson is all code. By the end, you will have working Python scripts that…
AI for Developers — Build with LLMs, RAG & Agents
You don’t need a PhD in machine learning to build with AI. You need to know how…
Building an AI Voicebot for Visitor Check-In — A Practical Guide to Handling the Messy Parts
Every office lobby has the same problem: a visitor walks in, nobody’s at the…