DeepSeek & Generative AI Masterclass for Software engineers

Name: DeepSeek & Generative AI Masterclass for Software engineers
Rating: 5.0 (9 reviews)

Learn AI/ML, Deep Learning, Generative AI & DeepSeek Architecture and internals with hands-on prompting and API access

Created byMohammed Fahim | Generative AI, Chatgpt trainer | AI professional

Last updated 6/2025

English

English [Auto],

What you'll learn

Artificial Intelligence : Basics of AI. Key Concepts of AI, AI/ML Differences, Deep Learning. Real life examples of AI, ML, Deep Learning and Generative AI.
Generative AI Internals : GPT (Generative Pre-trained Transformer), Gen AI timeline, Democratization of AI
DeepSeek : Effective Prompts to improve productivity. Build career - Create resume and prepare for interview. 1000+ Prompts with practical examples
DeepSeek Internals and DeepSeek Architecture
DeepSeek Benchmark comparison, Multi-head Latent Attention (MLA), Mixture-of-Experts (MoE), auxiliary-loss-free strategy, and multi-token prediction
DeepSeek Post training : Reinforcement Learning, Supervised learning, Chain-of-Thought (CoT), Distillation
DeepSeek : Programmatically access DeepSeek APIs, with AWS Bedrock, LM Studio

Course content

5 sections • 40 lectures • 3h 11m total length

Introduction to Artificial Intelligence and it's real-life applications2:54
Discover how artificial intelligence enables machines to learn from data, recognize patterns, and make data-driven decisions, illustrated by Netflix recommendations, Alexa, Google search, and self-driving cars.
AI, ML, Deep Learning & Generative AI comparison and examples4:08
Clarify the differences between artificial intelligence, machine learning, deep learning, and generative AI, showing how each term copies, learns, or generates content.
Artificial Intelligence(AI) v/s Machine Learning (ML)2:15
Learn the key differences between AI and ML, including how AI copies human intelligence and uses logic and rules, while ML learns from data to identify patterns and make predictions.
Typical ML Model building2:38
Explore how algorithms analyze past data, select features, and iteratively adjust parameters during training to produce accurate predictions on unseen data.
Understand Deep Learning with interesting example2:18
Generative AI - Power of Gen AI0:54
Explore the power of generative AI, which processes vast internet data with transformer models to understand and create text, images, and code.
Traditional AI v/s Generative AI1:42
Contrast traditional ai and generative ai across content creation, response quality, cost, and tasks, noting traditional ai excels at logical reasoning while generative ai favors creativity.
GPT (Generative Pre-Trained Transformer) - Understand properties of each keyword2:49
Explore how GPT processes input as token sequences with a transformer, enabling translation, summarization, and QnA, and how pre-training and generative capabilities adapt to images and videos.
Generative AI use cases & Limitations4:35
Explore generative AI use cases like text summarization, information extraction, translation, and code generation, alongside limitations like short-term memory, context retention, and hallucinations.
Generative AI Timeline - How Gen AI has evolved and role of Google and AWS2:29
Explore the evolution of generative AI from transformer innovations and open source breakthroughs to GPT three and ChatGPT, highlighting Google's and AWS's roles with bedrock and foundation models.
AI & Generative AI Introduction

Reinforcement Learning & Supervised Learning6:07
Explore supervised, unsupervised, reinforcement, and self-supervised learning, highlighting labeled data, pattern discovery, and automatic labeling with examples from chess and large language models.
What is Token and how to easily calculate tokens using tokenizer4:02
Explore tokens and tokenization in generative ai, learn how a tokenizer counts tokens for input and output, and understand billing implications across GPT models with practical examples.
Understanding Tokens and Temperature3:24
Explore how temperature controls randomness in generative AI, guiding predictable versus creative outputs, and how tokens and tokenization convert text to numbers for modeling.
Embeddings4:05
Explore embeddings as numeric codes that turn words or items into unique identifiers, reveal relationships, and power better predictions, search, recommendations, and language understanding.
Foundational Model6:51
Foundation models offer a single, versatile base trained on massive self-supervised data to tackle diverse tasks, with prompting and fine-tuning enabling low-label performance.
Foundation models, Gen AI & LLMs1:27
Distinguish foundation models, generative AI, and LLMs, and explore how each enables task adaptation, content generation, and language-focused text processing.
Generative AI Model training - Compare Fine Tuning, RAG and Few shot learning2:48
Explore how generative AI model training uses fine tuning to specialize on a task, RAG to pull knowledge, and few shot learning to generalize from few examples.
LLM Implementation options - APIs, Open Source LLMs and Fine tuning LLMs5:33
Compare LLM implementation options: commercial APIs, open source models hosted in your environment, and open source models trained on your data via fine-tuning; assess control, cost, and effort.
Transformer Architecture9:07
Explore the transformer architecture, its self-attention mechanism with queries, keys, values, and multi-head attention, plus positional embeddings, enabling long-term language understanding and superior translation.
Generative AI - Advance Topics

DeepSeek Login and initial screen introduction3:02
Log in to Deep Sea and explore the initial screen with chat history and prompts; compare free reasoning and web search options with ChatGPT.
DeepSeek Level 0 to Level 6 introduction2:31
Explore deep sea prompts from level zero to level six, mastering shell commands, Python programs, technical documentation, and end-to-end SDLC tasks like AWS architecture, Terraform, QA, and resume prep.
DeepSeek Prompts - Level 0 : Basics of Prompts & DeepSeek limitations5:26
Explore level zero prompts to master basics and deselect unnecessary reasoning or web search. Recognize limits like knowledge cutoffs and hallucinations; use 'I don't know' and internet access for info.
DeepSeek Prompts - Level 1 : Basic Shell commands4:37
Explore level one of Deep Sea Prompts to generate shell commands, including listing, copying, and removing files. Learn ls, find, cp, mkdir, and rm with practical flags.
DeepSeek Prompts - Level 2 : Advance Shell commands5:25
Master advanced shell commands by combining find and grep to search from root for credit or card mentions, handle sudo cautions, and tailor searches with prune and size filters.
DeepSeek Prompts - Level 3 : Basic coding and testing7:08
DeepSeek Prompts - Level 4 : Advance coding exercise8:51
DeepSeek prompts level four demonstrates building a recursive web scraper in JavaScript, comparing puppeteer and node-fetch, with robots.txt respect, rate limiting, and unit tests plus technical documentation.
DeepSeek Prompts - Level 5 : End to End project10:50
DeepSeek Prompts - Level 6 : Resume building and Interview preparation6:38
Advance your software engineering career with DeepSeek & Generative AI Masterclass level six, focusing on resume building, interview preparation, and showcasing big data, machine learning, and AWS skills.

DeepSeek Benchmark comparison6:56
Analyze how deep seq performs against OpenAI models across Amy 2024, Codeforces, gp q diamond, math 500, MLU, and SWE benchmarks, with reinforcement learning insights and cost advantages.
DeepSeek Internals - Introduction with interesting example6:27
Explore the internals of a deep sea generative AI model using a newsroom analogy, covering mixture of experts, auxiliary loss free strategy, multi-head latent attention, chain of thoughts, and distillation.
DeepSeek Architecture - Introduction, DeepSeekMOE & Auxiliary-loss-free strategy6:48
Explore the DeepSeek architecture, including DeepSeek MOE and MLA, and learn how the auxiliary loss free load balancing strategy and multi token prediction boost training efficiency and model performance.
DeepSeek Architecture -MLA (Multi-head Latent Attention)7:39
Explore DeepSeek architecture with multi-head latent attention (mla), featuring low-rank key-value joint compression, compressed latent queries and keys/values, rope rotary position embedding, and caching that improves inference efficiency.
DeepSeek Architecture -Multi-Token Prediction (MTP) & Distillation5:48
Explore DeepSeek's multi token prediction with a shared trunk and heads to speed up inference and generate text, and apply distillation from teacher to student models for efficient deployment.
DeepSeek Architecture - CoT, Group Relative Policy Optimization(GRPO) and FP85:16
Explore chain of thought reasoning to improve multi-step problem solving in ai models. Learn GRPO, group relative policy optimization, and FP8 quantization to boost reinforcement learning efficiency.
DeepSeek Internals and Architecture

DeepSeek API access2:13
Learn to programmatically access DeepSeek using an API key from platform.com and copy the key to enable API access, while noting a mid-February 2025 warning that API recharges are suspended.
DeepSeek API access via Openrouter.ai5:12
Access DeepSeek APIs via Open Router on openrouter.ai to choose from 300+ models and providers, learn uptime-based routing and API key creation for production workloads.
DeepSeek Python program5:55
Learn to build a Python program with deep seq APIs via Open Router and OpenAI libraries, using a free model, and a sample binary-tree depth calculation.
DeepSeek access through lmstudio.ai4:27
Discover accessing open-source models via lmstudio.ai, including Deepsee and Llama models. Learn downloading, loading, and running prompts such as binary search in Kotlin, with size and resource considerations.
DeepSeek access options in Amazon Bedrock5:25
Explore access options for the DeepSeek model on Amazon Bedrock, including serverless and marketplace deployments, model distinctions, and cost considerations for hosting.
DeepSeek API cost comparison with OpenAI ChatGPT reasoning models4:31
Compare deep sea api cost with OpenAI ChatGPT reasoning models to reveal pricing gaps, including input and output token costs, chain-of-thought accounting, and free chat options.

Requirements

Basic software engineering concepts

Description

Unlock the full potential of DeepSeek with our comprehensive course, designed to guide you from foundational concepts to advanced applications in Artificial Intelligence (AI), Machine Learning (ML), Deep Learning, and Generative AI.

Course Overview:

Foundations of AI and ML: Begin your journey by understanding the core principles of AI and ML, setting a solid groundwork for more complex topics.
Deep Learning and Generative AI: Delve into the intricacies of deep learning, exploring neural networks and their applications. Uncover the evolution and mechanics of generative AI, understanding how models like DeepSeek are revolutionizing content creation and problem-solving.
Advanced AI Topics: Expand your knowledge with advanced subjects such as Reinforcement Learning, Tokenization, Temperature Scaling, Embeddings, and Foundation Models. Gain insights into the training methodologies of generative AI models and dissect the Transformer architecture that powers them.
Hands-On DeepSeek Prompting: Transition from theory to practice with hands-on sessions focused on DeepSeek prompting. Learn to craft prompts ranging from basic to advanced levels, enhancing your skills in software development life cycle (SDLC) processes and career advancement.
DeepSeek Internals and Architecture: Explore the technical marvel that is DeepSeek. Study its architecture in detail, including innovative techniques like Mixture of Experts, Multi-Head Latent Attention, Multi-Token Prediction, Distillation, Chain-of-Thought (CoT), Group Relative Policy Optimization (GRPO), and FP8. These lectures are enriched with insights from leading AI research papers.
Programmatic Access to DeepSeek: Learn how to integrate DeepSeek into your projects programmatically. Gain proficiency in accessing DeepSeek through APIs, utilizing platforms like Amazon Bedrock and LM studio. Engage with practical examples and compare API costs with services like OpenAI's ChatGPT, understanding DeepSeek's cost-effective advantages.

Why Choose This Course?

Comprehensive Curriculum: From basics to advanced topics, this course covers all essential aspects of DeepSeek and its applications in AI.
Practical Application: Engage in hands-on exercises that bridge the gap between theoretical knowledge and real-world application.
Expert Insights: Learn from content derived from cutting-edge AI research, ensuring you receive up-to-date and relevant information.
Cost Efficiency: Understand how DeepSeek offers a competitive edge with its cost-effective solutions compared to other AI models.

Don't miss this opportunity to elevate your understanding of DeepSeek and Generative AI. Enroll now to unlock a world of hands-on examples, in-depth architectural insights, and practical knowledge on implementing this groundbreaking AI technology.

Who this course is for:

Software Engineers
QA Engineers
Software Professionals
AI/ML Engineers
Data Scientists

DeepSeek & Generative AI Masterclass for Software engineers

What you'll learn

Explore related topics

Course content

Intro to AI/ML & Generative AI - Watch at 1.5x if you already know these well10 lectures • 27min

Advance Topics - Reinforcement Learning, Tokens, LLMs, Embedding, Transformer9 lectures • 43min

DeepSeek Hands-on : Basics of Prompts to End 2 End Project and Resume building9 lectures • 54min

DeepSeek Internals - Benchmark, Architecture and explanation of each component6 lectures • 39min

DeepSeek Model access through API, Amazon Bedrock, lmstudio.ai and Hands-on6 lectures • 28min

Requirements

Description

Who this course is for: