Precious Oaseru Osaro is an engineering leader who enjoys building products, solving hard problems, and developing high-performing teams that solve problems. He is currently building projects like JustRever—an AI-powered scam protection platform and Memory, a product designed to help our memories live forever. Precious explores the future of engineering leadership, software development and team culture. And also loves to mentor and help others learn and grow.
Take the Agentic Engineering Audit and discover your level across 10 categories of AI-powered engineering mastery.
Take the AuditPart 3 of the Edge AI on iOS series. Build a complete on-device RAG pipeline in Swift with no dependencies: NLContextualEmbedding for semantic search, Foundation Models for generation, and the surprising thing that happens when retrieval is a tool the model calls.
Part 1 of the Edge AI on iOS series. Roughly two billion phones now ship with a usable on-device language model. A working overview of what is actually shipping, what people build with it, and where on-device intelligence is heading.
Part 2 of the Edge AI on iOS series. The bottleneck is not compute, it is memory bandwidth. What on-device AI actually costs in battery, heat, and accuracy, what the next silicon generation changes, and what to build now.
An open-source AI agent skill that scans your machine and project for agentic engineering maturity. 9 categories, 0-900 points, 10 levels. Download it and audit yourself.
A unified framework for measuring your AI agent maturity, from terminal tourist to swarm architect. Find out where you stand on the 0-900 scale.
Why 'You are a senior developer...' prompts are becoming irrelevant, and what's replacing them. The shift from prompt crafting to context design.
Every major model compared by role, cost tier, and use case. Plus the routing pattern that cuts costs by 65%.
A practical guide to setting up Claude Code with CLAUDE.md, custom skills, and MCP servers.
The same three-layer framework applied to GitHub Copilot's ecosystem. Context, skills, and tools compared to Claude Code.
Cursor-specific configuration for context engineering with .cursorrules, plan mode, and multi-agent.
Plan, Delegate, Assess, Codify. How AI agents improve with every session through structured feedback.
PreSession, PostTask, PostToolUse, PostApproval: how hooks make everything automatic.
Sequential, parallel, multi-model, feedback loops. Plus the reviewer rule and Gas Town.
The Model Context Protocol explained: what it is, how to set it up, and why it is the USB-C for AI.
Don't chase tools. Analyse YOUR workflow. Map AI into YOUR process. Build from there.
Happy path, error path, personas, edge cases. What agile skipped and AI needs.
How to write specs that AI agents can execute precisely. The planner agent workflow.
The right to say "this is not ready", one of the most important rights a developer has, yet the one that has lost its impact the most, especially when faced with unfinished requirements, unclear product requirement and intense management pressure
Complete technical guide to passkeys, OPAQUE protocol, and cryptographic authentication. Learn WebAuthn, FIDO2 implementation, authentication lifecycle, and security best practices for iOS and Android.
After nearly 1,500 minutes of mentoring engineers at every level, I finally understand what separates great engineers from those who just write code.