CSE453 — MLOps and LLMOps: Engineering AI Systems | IBA Karachi (Fall 2025)

A structured course repository for MLOps and LLMOps: Engineering AI Systems at Institute of Business Administration (IBA), Karachi.
This repo organizes lecture material, study notes, midterm prep packs, and revision resources in a clean week-and-topic friendly layout.

Maintained by: Zuha Aqib (Student, IBA Karachi)

---

Course Information

• Institute: Institute of Business Administration (IBA), Karachi
• School: School of Mathematics and Computer Science
• Semester: Fall 2025
• Course Title: MLOps and LLMOps: Engineering AI Systems
• Instructor: Sualeh Ali
• Credits: 3 (2-1)

---

Table of Contents

• Quick Links
• What’s Inside
• Repository Structure
• Course Topic Map
• Milestones & Assessment
• How to Use This Repo
  • Midterm Preparation
  • After Mid / Final Preparation
• Notes & Disclaimer

---

Quick Links

• Course Outline (PDF)
• Course Outline (Duplicate Copy)
• Lectures
• Class Notes
• Guest Speaker Sessions
• Mid Exam Pack
• Revision Notes (After Mid)
• Reference Book

---

What’s Inside

• Course outline PDFs
• Lectures (PDF/PPTX + a tutorial notebook)
• Handwritten class notes (before and after mid)
• Guest speaker session notes
• Mid exam pack
  • Topic-wise PDFs (lecture-by-lecture)
  • Mid topics list
  • Mock exams
  • Revision sets
• Revision notes after mid
  • LLM intro, pretraining
  • RAG
  • Mixed MCQs
• Reference book (Practical MLOps)

---

Repository Structure

.
├── Deep_Learning_Course_Outline_(2).pdf
├── MLOpsLLMOpsCourseOutline.pdf
├── book/
│   └── Practical_mlops_-_Noah_Gift.pdf
├── class_notes/
│   ├── after_mid.pdf
│   ├── before_mid.pdf
│   └── WhyMLOps_notes.pdf
├── guest_speaker_sessions/
│   └── Meeting_started_-_Notes_by_Gemini_(1).pdf
├── lectures/
│   ├── What_is_MLOps_anyways.pdf
│   ├── ML_Toolbox__1_.pdf
│   ├── AutoML_and_such__1_.pdf
│   ├── ML_Observabilitty_with_Evidently_AI__1_.pdf
│   ├── MLOps_LLMOps_Cloud_Comparison__1_.pptx
│   ├── RAGs_on_RAGs_on_RAGs.pptx
│   ├── A_Crash_Course_of_the_‘LLM’.pptx
│   ├── LLMS__Alignment,_Evaluation.pdf
│   ├── Milestone_1.pdf
│   └── llmops_comprehensive_tutorial.ipynb
├── mid_exam/
│   ├── mid_topics.txt
│   ├── lec01_what-is-mlops.pdf
│   ├── lec02_virtual-env.pdf
│   ├── lec03_containers-and-vm.pdf
│   ├── lec04_docker.pdf
│   ├── lec05_kubernetes.pdf
│   ├── lec06_feature-stores.pdf
│   ├── lec07_automl.pdf
│   ├── lec08_cloud.pdf
│   ├── lec09_mlflow.pdf
│   ├── lec10_drift.pdf
│   ├── lec11_ci-cd-fastapi.pdf
│   ├── lec12_conda.pdf
│   ├── mock-exam-1.pdf
│   ├── mock-exam-2.pdf
│   ├── mock-exam-3.pdf
│   ├── mock-exam-4.pdf
│   ├── revision-1.pdf
│   └── revision-2.pdf
└── revision_notes/
    ├── llm-intro.pdf
    ├── llm-pretraining.pdf
    ├── rag.pdf
    └── mixed-mcqs.pdf

---

Course Topic Map

High-level map of what the course covers (MLOps first, then LLMOps):

1. MLOps lifecycle: from notebooks to production systems
2. Git, environments (venv/conda), and maintainable project structure
3. Docker and serving ML via APIs (FastAPI)
4. CI/CD and automation (e.g., GitHub Actions)
5. Monitoring traditional ML: drift, metrics, dashboards (e.g., Evidently)
6. The shift to LLMOps: evaluation challenges, prompt engineering, RAG
7. Building LLM applications with RAG + vector databases
8. Evaluating LLM/RAG systems (RAG triad, AI-as-judge, evaluation frameworks)
9. CI/CD for LLMs: quality gates + prompt versioning
10. Monitoring LLMs: cost, latency, quality, observability tools
11. Responsible AI & security: bias, toxicity, prompt injection, governance

---

Milestones & Assessment

The course is milestone/project-driven:

• Milestone 0: Project pitch (proposal + short presentation)
• Milestone 1: Traditional ML tool deployment with monitoring (containerized + API + observability)
• Milestone 2: LLM component deployment (RAG + vector DB + evaluation)
• Final: End-to-end system presentation + report (architecture, deployment, monitoring, evaluation, ethics)

Assessment (as per outline):

• Class participation
• Quizzes
• Midterm
• Final milestone-based project (Milestone 0/1/2 + final deliverable)

---

How to Use This Repo

Midterm Preparation

Use the midterm pack in this order:

1. Mid topics list: mid_exam/mid_topics.txt

2. Topic-wise notes: mid_exam/lec01_... to mid_exam/lec12_...

3. Revision sets: mid_exam/revision-1.pdf, mid_exam/revision-2.pdf

4. Mock exams: mid_exam/mock-exam-1.pdf to mid_exam/mock-exam-4.pdf

5. Cross-check weak areas using:

   • lectures/ (core lecture files)
   • class_notes/before_mid.pdf

After Mid / Final Preparation

please note: there was no final exam for this course.

1. Start from class_notes/after_mid.pdf to align with in-class direction.

2. Use the after-mid notes in revision_notes/:

   • llm-intro.pdf
   • llm-pretraining.pdf
   • rag.pdf
   • mixed-mcqs.pdf

3. Pair with the relevant lecture files in lectures/ (LLM/RAG/observability content).

4. Use the book in book/ for deeper understanding and real-world framing.

---

Notes & Disclaimer

• This repository is intended for learning and revision.
• Lecture material and outlines belong to their respective authors/instructors.
• Any self-made mock exams/notes are included for practice and study support.