⤵️ Quick shortcut: Select the module you’re currently studying ⤵️

  Courses      Module 1     Module 2      Module 3     Module 4     Module 5     Module 6    

This curriculum is a hands-on guide for senior software engineers to learn how to build a complete inference pipeline using LLMs. It walks you through key steps like data ingestion, preprocessing, embeddings, retrieval-augmented generation (RAG), and more. These aren’t just nice-to-have skills anymore—they’re becoming essential, like knowing Git or deploying to AWS. Sure, calling a ChatGPT API might seem like enough at first, but if you stop there, you’re leaving performance, cost savings, and better results on the table. Building the whole pipeline means you can optimize for your specific use case, fine-tune for better answers, and run things more efficiently. This curriculum cuts through the noise and shows you exactly how to do it, step by step.

The objective of this course is to equip learners with the skills to build a production-ready pipeline for leveraging Large Language Models (LLMs), an increasingly critical component of modern software development. LLM-powered features are becoming standard in various industries, and our partners expect this expertise as part of a developer's toolkit.

This course is project-based and revolves around building Jarvid, a practical example of a production-grade LLM-powered pipeline.

• What is Jarvid?
  Jarvid is a system designed to address a common business challenge: connecting users with relevant expertise and knowledge from a large database efficiently and intelligently.

  1. Demo Components: Include a video demonstration, link to a working prototype, or a text walkthrough.

  2. Key Business Problem Solved: Automating expert recommendations by retrieving contextually relevant employee data.

• Major Architectural and Pipeline Components:

  1. Data Ingest: Collecting raw input data.

  2. Pre-processing: Cleaning, formatting, and preparing data for use.

  3. Embedding: Representing text as vectorized data for retrieval tasks.

  4. RAG (Retrieval-Augmented Generation): Integrating retrieved data with LLM-generated responses.

  5. Prompt Design: Crafting LLM prompts for effective responses.

  6. Deployment: Building scalable and reliable APIs.

By emphasizing these modular components, learners will see how the techniques extend across domains and tasks beyond Jarvid.