
How We Actually Use AI to Ship Code
A practical look at how we direct coding agents, protect context, debug with evidence, test generated code, and keep human judgment in the loop.
Topic
AI engineering is the work between a cool demo and something a team can trust. This topic collects roadmaps, architecture questions, evals, deployment lessons, and hiring advice.

A practical look at how we direct coding agents, protect context, debug with evidence, test generated code, and keep human judgment in the loop.

We measured a dozen context management strategies on our production AI tutor, first on cached cloud models and then on local ones, and the field's default advice inverted: under modern prompt caching, keeping everything beat every summarization strategy on cost, latency, and memory at once.

I am excited to launch the mentorship I wish every solo AI engineer could have: a senior team behind your projects, decisions, and career.

A practical path from coding-agent speed to the architecture, evaluation, and failure-analysis skills real AI engineering work requires.

The architecture, agent loops, writing workflow, visual system, and title tests behind a content pipeline that cut my workload from two or three days to about five hours.

Why the loops-versus-graphs debate misses the point, what actually changed in agent orchestration, and which parts of your system should stay deterministic.

A practical way to stop babysitting coding agents: design their context, tools, checks, budgets, and stop conditions.

What changed in AI coding, why generated code can feel faster, and where engineering judgment still matters.

How recursive AI improvement loops work, why weak evals reward the wrong behavior, and what builders can use now.

Prompting tells an agent what to do. The harness controls its tools, permissions, tests, traces, and failures.

After interviewing more than 100 candidates, here’s what actually stands out in AI engineering interviews and take-home assignments

What separates a polished AI agent demo from a system your team can evaluate, debug, deploy, and trust.

A cheatsheet to avoid costly rework in agent systems.

You’re not building agents. You’re building workflows (and that’s fine)

1 referral = our bestselling book. 3 referrals = a course. 10 = full access + affiliate tier


Tips for Crafting the Perfect Prompt for Each Model

(full training session) (typical path for companies)


How Open-Sora 2.0 Built Sora-Level Video AI for $200K

The Need for New Skills and Roles


Research and Writing: Why AI Tools Change the Workflow

API Deployment for AI Engineers

Optimizing Large Language Models for Retrieval-Augmented Generation

Is fine-tuning an embedding model worth it?

Retrieval-Augmented Generation vs. Long Context: A Comprehensive Comparison

What Makes AI Truly Useful?

Why RAG Evaluation Matters and Techniques to Leverage

Advanced Vector Indexing Techniques for RAG

No-code Custom LLM Evaluation Demo




Our first book: Building LLMs for Production: Enhancing LLM Abilities and Reliability with Prompting, Fine-Tuning, and RAG

The What's AI podcast episode 28 with artist and CEO Mariam Brian

Jerome Pasquero on Data Annotation: The Building Blocks of AI's Future - The What's AI Podcast (episode 27)

AI, autonomous vehicles and more in this new episode of the What's AI podcast!

The What's AI podcast episode 25 with Jerry Liu: LlamaIndex CEO and co-founder

Tips on what to do with your language model or API

How we Built an Open-Source RAG-based ChatGPT Web App

The What's AI Podcast Episode 20 with Auxane Boch: Ethics Expert

Move anything from point A to point B!

Interview with Yotam Azriel, CTO at TensorLeap - What's AI episode 13

An interview with Limarc Ambalina, VP of Editorial at HackerNoon

Ph.D., Twitter, Breaking into the AI Field, and more with Brian Burns (AI Pub)

These AI Tools will 10x Your Productivity!

Insights to become a self-driving car engineer in 2023


An interview with the Director of Perception at Zoox, Ruijie (RJ) He, with the goal of demystifying what is a good profile to get an ML engineer job and perform at the interviews.

Disney's New Model Explained

What is a prompt engineer and how to improve at it…

What does such a model understand when it sees such a picture or, even more complex, a video?

A High-Resolution Image Synthesis Architecture: Latent Diffusion

The beginning of data-centric AI with data programming

How can Uber deliver food and always arrive on time or a few minutes before?

the five best articles related to artificial intelligence in February

AI-based face manipulations in high-quality talking head videos !

"Sleep is not a disposable luxury - it's a non-negotiable biological necessity." -Matthew Walker

Imitating the nematode's nervous system to process information efficiently, this new intelligent system is more robust, more interpretable, and faster to train than current deep neural network architectures with millions of parameters.

Andrej Karpathy's talk on Tesla's autopilot explained clearly in under 10 minutes

Say goodbye to complex GAN and transformer architectures for image generation. This new method can generate new images from any user-based inputs.

AI is everywhere; understand how it works and build the most powerful applications! A simple explanation of the most popular deep neural network architecture for images: Convolutional Neural Networks. AI put simply.

Using a modified GAN architecture, they can move objects in the image without affecting the background or the other objects!

This AI can be applied live to the video game and transform every frame to look much more natural.

Properly relight any portrait based on the lighting of the new background you add.

With this AI-powered nerve interface, the amputee can control a neuroprosthetic hand with life-like dexterity and intuitiveness.

In less than 5 minutes, you will know how the transformer architecture can be applied to computer vision with a new paper called the Swin Transformer

Drawing inspiration from Human Capabilities Towards a more general and trustworthy AI & 10 Questions for the AI Research Community.

A professional data scientist answers the 12 most asked questions. Clarify and debunks the myths you've built surrounding this field!

This method is called DeOldify and works on pretty much any picture. If you don’t believe me, you can even try it yourself for free, as I will show in this article.

This artificial intelligence model transfers clothes between photos using only one picture of yourself and one picture of the clothing! This is called garment transfer...

This AI can transform any of your pictures into an accurate representation with a Disney animated movie character style!

Understand how deep fakes work and create your own!

YOLO, Also Known as You Only Look Once is one of the most powerful real-time object detector algorithms. It is called that way because…
Choose whether Google Analytics can use cookies. The site always keeps anonymous, cookieless page counts. Privacy details