Skip to main content
Back to CoursesCareer TrackSoftware Engineering

AI Engineering & MLOps Course in Nepal — LLMs, Generative AI, Model Deployment & Production ML

Build production AI systems with RAG, agent workflows, evaluation, and MLOps discipline.

Duration
12 weeks
Batch Size
Max 10
Format
Online + Offline
Level
advanced

Your Learning Schedule

Mon-Fri live classes

2 hrs live class/day + 2 hrs self-study at home (required).

Sunday Open Lab

Classrooms and labs stay fully open all day. Come study, pair-program, and build.

Daily commitment

Minimum 2 hrs focused practice beyond class at home. This is what builds real mastery.

Core tech & tools
PythonLangChainLangGraphHugging FaceStreamlitRAGRetrieval EvaluationMLflow

Production AI thinking: structured outputs, retries, guardrails, and cleaner service boundaries

RAG systems with stronger retrieval, chunking, evaluation, and citation habits

Agent workflows with LangGraph and multi-step orchestration patterns

Tuition & Support
Initial batch offer

Mar 1, 2026 – Dec 31, 2099

NPR 40,000Flexible options available
NPR 34,000

Save NPR 6,000 on this course right now.

Live interactive sessions with trainers

Small-batch feedback on your practical work

Course materials, recordings, and lab support

Personalized guidance after Saarathi Gate on pacing, practice focus, and support

Upcoming Batch

Enrolling Now

Limited seats. Final class timing is confirmed within 48 hours based on availability.

About this Course

This course is built for learners who already have data and ML foundations and now need production AI engineering discipline. The focus is not only on using LLM APIs, but on structuring outputs, improving retrieval quality, evaluating responses, and building systems other people can review and maintain.

Students move from notebook-level experimentation into more dependable AI workflows: RAG systems, tool-using agents, evaluation pipelines, local inference awareness, and MLOps habits that make deployment less fragile.

Production AI thinking: structured outputs, retries, guardrails, and cleaner service boundaries

RAG systems with stronger retrieval, chunking, evaluation, and citation habits

Agent workflows with LangGraph and multi-step orchestration patterns

Fine-tuning and local inference taught with judgment instead of hype

MLOps delivery with MLflow, FastAPI, monitoring, and trainer-reviewed capstone work

Capstone spotlight

AI-Powered Document Intelligence Platform

Final build focus

> Note: Project themes may evolve each batch, but every learner still completes one trainer-approved final project from the approved pool. Smaller guided exercises can happen during the course, but the public completion standard stays anchored to one polished final outcome.

Likely stack

PythonLangGraphPineconeHugging FaceMLflowFastAPIStreamlit
  • Multi-format document ingestion pipeline (PDF, DOCX, web)
  • Advanced RAG with re-ranking and citation tracking
  • AI agent for complex multi-step document analysis
  • Fine-tuned model for domain-specific extraction and evaluation

Why This Course?

Global Context: Teams are moving from AI demos to production systems that need clearer retrieval, evaluation, guardrails, and monitoring. The engineering challenge is no longer just calling a model. It is making the system dependable.

Nepal Context: Nepal's AI ecosystem continues to grow through product companies, service teams, research organizations, and remote engineering work. The gap is not only model access. It is in engineers who can turn AI ideas into maintainable services with testing, monitoring, and better cost awareness.

Your Opportunity: This course is built for the learner who already has ML basics and now wants stronger production judgment. You will practice RAG, agents, evaluation, local inference awareness, and MLOps in a way that rewards reliability over hype.

The strongest AI engineers are usually the ones who can explain why a system retrieved the wrong context, why a prompt failed, why a model should not be fine-tuned yet, and how to monitor quality after launch.

Reliable retrieval, evaluation, and monitoring habits matter more than demo polish

Teams are moving from AI experiments to production services

Structured outputs, retries, and system boundaries become core engineering work

Prompting alone stops being enough in production

Knowing when not to fine-tune is part of good judgment

Fine-tuning is accessible but often overused

Versioning, evaluation, and rollback discipline create real team value

MLOps gaps slow down AI delivery

How Saarathi Gate shapes your learning plan

Saarathi Gate is a diagnostic, not a pass-or-fail exam. It helps us understand your current skill level, how you learn best, where you are already strong, and where you need extra support before the batch begins.

Before the batch starts

You complete Saarathi Gate so we can understand your current skill level, how you learn best, your strengths, and the support you may need before classes begin.

During the course

Trainers use that diagnostic profile to guide pacing, practice focus, feedback, and the kind of support that helps you learn best.

Certification and proof of work

Certification for AI Engineering & MLOps Course in Nepal — LLMs, Generative AI, Model Deployment & Production ML depends on attendance, required coursework, trainer review, and the practical work described in the micro-syllabus and full syllabus.

Curriculum

A structured journey to mastery

WeekFocus AreaWhat You'll Master
W1Production AI Foundations & Structured OutputsL1 recap, production AI mindset, prompt discipline, structured outputs, provider APIs, and the difference between a demo and a dependable service
W2Vector Databases & Retrieval QualityEmbeddings, vector stores, chunking, metadata design, indexing strategies, and retrieval-quality trade-offs
W3RAG Pipeline Development & EvaluationRAG architecture, retrieval optimization, context assembly, citation-aware responses, and evaluation basics

Where this course can lead

These are examples of roles, responsibilities, or directions this course can help you grow toward.

Possible next roles
Junior AI Engineer / Applied ML Engineer
AI Engineer / MLOps Engineer
Senior AI Engineer / ML Platform Engineer
AI Solutions Architect / Applied AI Lead

Frequently Asked Questions

Completion of Data Science & Machine Learning (TR-02) or equivalent is required. You should already be comfortable with Python, pandas, basic model training and evaluation, notebook workflow, and reading technical documentation. Before Day 1, review train/test split, leakage, prompt basics, and transformer fundamentals, then complete the AI readiness assessment. All applicants complete Saarathi Gate Assessment and an advanced readiness review before enrollment.

Seats are limited

Ready to start your journey?

Complete Saarathi Gate, confirm your level, and secure your seat for the upcoming batch.