Prompt Engineering & LLMs
You already use AI. Whether for work, personal tasks, travel planning, or something else, AI is becoming part of everyday life. But are you truly getting the most out of it? Have you ever felt disappointed or underwhelmed by the results an AI tool produced?
At Navalapp, we optimize ships and other complex systems (AI & Machine Learning: From Scratch to Advanced Models) —so why not optimize the prompts we give to generative AI models as well?
In this series, we’ll explore research-backed techniques (with Amatriain, X., “Prompt Design and Engineering: Introduction and Advanced Methods,” 2024, arXiv, as our main reference) and take a quick look under the hood of Large Language Models (LLMs) to help you turn “meh” outputs into trustworthy, effective, and actionable ones.
What’s in the series:
- What’s a prompt? + basic patterns
Understand the anatomy of an effective prompt and get three ready-to-use templates (Instructions + Question, Instructions + Data, Question + Examples).
- LLM foundations
A high-level look at how models read (tokens & context windows), think (next-token prediction), and create(Transformers). Learn why length, structure, and placement significantly influence results.
- Advanced prompt strategies
Chain-of-Thought & Manual CoT, Tree-of-Thought, Self-Consistency (sample & vote), Reflection (answer → critique → revise), Expert prompting, and Rails (topical, factual, safety) using simple JSON schemas.
- What LLMs are used for?
Information retrieval (with verification), writing & editing (edit-first prompt), practical planning (constraints that actually save time), technical troubleshooting (MRE triage prompts), and light multimodal use (tables, plots, images).
What you’ll get out of it:
- More reliable first drafts (less back-and-forth)
- Short, reusable prompt templates for everyday tasks
- A practical mental model of how LLMs behave—and how respond to structure
Join the discussion in the forum: ask questions or share a task you’d like to optimize.
Subscribe to the forum thread so you don’t miss any updates. It’s free!
