PIOAI

Navigate

Advanced Certificate in LLM Engineering

Build language systems that survive contact with real users.

Apply to this programTalk to admissions
Duration
16 weeks
Format
Online · Cohort-based, two live studios a week
Tuition
Rs 260,000
Intake
March and October cohorts

Overview

Most LLM courses end where the difficulty begins. This certificate starts there. You will build the same assistant three times — naïve, retrieval-augmented, agentic — and measure each against an evaluation set you construct, so that every architectural decision is defended with numbers.

Studios are live and small. Each week pairs a lecture on a concept (context engineering, structured outputs, guardrails) with a lab that breaks something on purpose.

Outcomes

  • Design context and retrieval strategies with measured relevance and cost.
  • Construct evaluation sets and LLM-as-judge pipelines you can trust.
  • Build tool-using agents with permission boundaries and observability.
  • Operate LLM features with budgets, caching and incident response.

Curriculum

  1. 01

    Weeks 1–4 · Language models as components

    • NLP with transformers (refresher)
    • Prompting, structured outputs and caching
    • Building the naïve baseline
  2. 02

    Weeks 5–10 · Retrieval and evaluation

    • LLM engineering and evaluation
    • Chunking, embeddings, hybrid search
    • Constructing an evaluation set
    • Project: measured RAG assistant
  3. 03

    Weeks 11–16 · Agents and operations

    • Tool use, planning and memory
    • Safety, guardrails and red-teaming
    • Cost and latency engineering
    • Capstone: agentic workflow with evals

Courses in this program

LanguageFeatured

LLM Engineering & Evaluation

Build language systems you can measure

4 modules · 11 lessons30h2 enrolled
Dr. Bilal Ahmed
Advanced
Open course