Agentic AI Masters 2026: LangChain, LangGraph & CrewAI

Explore map, reduce, and filter in Python using lambda expressions for concise, loop-free code. See practical examples like calculating circle areas and filtering data, with reduce noted as deprecated.

Explore control structures in Python, including binary and relational operators, assignment, boolean outcomes, and if-else decision making, then practice with inputs, range, lists, and map, filter, reduce.

Explore object oriented programming in Python by creating classes and objects, using constructors, and applying concepts like inheritance, encapsulation, and polymorphism.

Learn numpy, the python library for numerical operations on one to n dimensional arrays, and master creating, shaping, indexing, and performing arithmetic with zeros, ones, full, identity, and random arrays.

Uncover how Python data visualization translates numbers into insights with matplotlib and seaborn. Identify patterns, correlations, and trends using charts like bar, line, area, and pie to support data-driven decisions.

Explore how matplotlib and pandas enable plotting histograms, bar charts, pie charts, scatter plots, and more, using data frames and basic Python tooling.

Explore Seaborn, a powerful Python data visualization library, covering installation, using iris data, and creating kd plots, dist plots, pair plots, heat maps, and box plots.

Explore the history and challenges of natural language processing, and learn how NLP enables translation, sentiment analysis, and chatbots using large language models and LangChain.

Explore the key challenges in NLP, including pragmatic, lexical, syntactic, and anaphoric ambiguity. Assess issues of lack of standardization, ethical considerations, context understanding, and data sparsity in specialized domains.

Learn case folding in natural language processing, converting text to lowercase to normalize data and improve generalization, with Python examples using casefold and lower, and note limitations for proper nouns.

Master special character removal, a vital NLP pre-processing step that eliminates non-alphanumeric characters to reduce noise, improve tokenization, and enhance model output, demonstrated with re, spaCy, and NLTK.

Learn tokenization as a fundamental NLP step, breaking text into sentences and words with word, sentence, and subword tokenizers. Practice with Python and the NLTK library to illustrate tokenization concepts.

Identify stop words, remove them to boost signal-to-noise ratio in NLP text data, using Python code with NLTK stopwords and word tokenize for preprocessing.

Explore n-grams in NLP, including unigrams, bigrams, and trigrams, and learn how to build language models, apply to spelling correction, text summarization, and feature engineering with Python NLTK.

Learn how textual data becomes numerical vectors through vectorization, enabling models to process language; explore one-hot encoding, bag of words, tf-idf, and word embeddings like word2vec, glove, and fasttext.

Explains bag of words as a foundational vectorization method, turning text into binary word presence and vectors, while noting stopword removal and the practical drawbacks tf-idf later addresses.

Learn practical bag of words with a spam detection example using a labeled dataset, count vectorization, train/test split, and a Multinomial Naive Bayes classifier to evaluate performance.

Explore tf-idf, a statistic that blends term frequency and inverse document frequency to measure a term's importance and downplay common terms while guiding word vectorization.

Learn practical tf-idf concepts with a hands-on demo using a tf-idf vectorizer, term frequency and inverse document frequency for e-commerce text classification.

Master part of speech tagging and named entity recognition as NLP pre-processing, with practical Spacy demos and introductions to word sense disambiguation and Hidden Markov–Viterbi concepts.

Demonstrate named entity recognition with spaCy through practical examples, identifying entities like organizations and monetary values, examining labels and pipelines, and creating custom entities in a doc.

Explore pretrained word2vec vectors from Google News (300 dimensions) using gensim, mastering word similarity, analogy, and text classification with cosine similarity and vector arithmetic (king minus man plus woman).

The lecture offers an intuition for word2vec, showing king minus man plus woman producing queen, and explains automating features with neural networks and reducing 300 features to 50.

Examine how 300-dimensional Word2Vec vectors reduce to the first 50 features and visualize them with a heat map to compare words like king, queen, woman, girl, boy, and man.

Learn how Word2Vec CBOW converts words into vectors using a context window, neural networks, softmax, and backpropagation to predict target words from surrounding context.

Explore glove, global vectors, and how word-to-word co-occurrence probabilities generate vector representations through matrix factorization, enabling semantic relationships and cosine similarity insights.

Discover fastText as an upgraded word embeddings method from Facebook, faster than word2vec and capable of using character or n-gram units to handle out-of-vocabulary terms and non-English data.

Explore cosine similarity as the angle between non-zero vectors, independent of magnitude. See its use in nlp with sentences via bag-of-words and one-hot representations.