Codersarts Blog.

Course:  LLM Foundational Course Level:  Medium → Advanced Type:  Individual Assignment Duration:  5–7 days Total Marks:  100 Objective The objective of this assignment is to help you: Implement conversation memory  that manages context windows Control LLM output  using temperature, max_tokens, and stop sequences Build a complete agent  that combines conversation history with semantic search Handle multi-turn conversations  with proper context management Track usage and costs

Course:  LLM Foundational Course Level:  Medium Type:  Individual Assignment Duration:  5–7 days Total Marks:  100 Objective The objective of this assignment is to help you: Understand tokenization  and how it affects API costs Implement token counting  and cost calculation functions Build a vector database from scratch  using embeddings Perform semantic search  to retrieve relevant documents Create a simple RAG system  that answers questions using retrieved context Think pra

Course:  AI Backend Engineering with FastAPI  Assignment Type:  Applied Project Assignment  Difficulty Level:  Medium → Advanced  Estimated Time:  12–18 hours  Submission Mode:  Online LMS (Moodle / Canvas / Blackboard) Assignment Context Modern AI systems rarely consist of a single model endpoint. Instead, they involve multiple components working together such as: API orchestration model inference retrieval systems asynchronous processing background workflows logging and mon

Course:  AI Backend Engineering with FastAPI  Assignment Type:  Capstone Implementation + Architecture Report  Difficulty:  Medium → Advanced  Estimated Effort:  10–15 hours  Submission Platform:  LMS (Moodle / Canvas / Blackboard) Assignment Overview In this assignment, you will design and implement a production-style AI backend API  using the concepts introduced throughout this course. You will build a modular FastAPI system that integrates: LLM inference Retrieval-Augmente

Course: MCP Fundamentals  Level: Medium → Advanced Type: Individual Assignment Duration: 5–7 days Objective The objective of this assignment is to help you: Understand advanced agentic MCP capabilities (Sampling, Elicitation, Roots) Design multi-agent systems with appropriate orchestration patterns Analyze security implications of agentic workflows Implement human-in-the-loop design patterns Reason about long-running workflows and error handling Think critically about prod

Course: MCP Fundamentals  Level: Medium Type: Individual Assignment Duration: 4–5 days Objective The objective of this assignment is to help you: Understand the architectural problem MCP solves and why earlier approaches failed Master the distinction between Tools, Resources, and Prompts Apply primitive selection logic to real-world integration scenarios Design MCP Server architectures with appropriate primitives Analyze trade-offs in transport mechanisms and deployment m

Objective In this assignment, you will move beyond isolated chunking techniques and design a complete, adaptive chunking system  that intelligently selects or combines strategies based on the input document type. This is closer to how chunking is actually used in production systems. Problem Statement Most tutorials treat chunking strategies independently: Fixed-size chunking Overlapping chunking Sentence-based chunking Token-aware chunking Semantic chunking However, in real-w

Course: Chunking Strategies for Production RAG Systems Level: Medium → Advanced Type: Individual Assignment Duration: 5–7 days Objective The objective of this assignment is to help you: Understand and implement multiple chunking strategies Analyze trade-offs between different approaches Design a hybrid chunking pipeline Evaluate chunking quality in a Retrieval-Augmented Generation (RAG) context Think like an engineer building production-ready systems Problem Statement You a

Course: RAG Evaluation Level: Medium to Advanced Type: Individual Duration: 7 to 10 days Objective This assignment tests your ability to evaluate the generation stage of a RAG pipeline, attribute failures to the correct pipeline stage, and automate the entire evaluation workflow using an LLM as a judge. You will generate RAG answers, measure faithfulness and completeness, run end-to-end error attribution, build a structured LLM judge, and compare automated scores against your

Course: RAG from Scratch Level: Beginner to Medium Type: Individual Duration: 5 to 7 days Objective This assignment tests your ability to build the foundational stages of a RAG pipeline: loading documents, extracting clean text, attaching metadata, enriching documents with LLM-generated keywords, and splitting them into retrievable chunks. By completing this assignment, you will have built a reusable knowledge base preparation pipeline that you can apply to any document colle

Introduction Modern satellite constellations generate petabytes of multispectral, hyperspectral, SAR, and LiDAR data every day, far outpacing the capacity of traditional analysis methods. Remote sensing professionals must interpret this imagery against historical baselines, evolving scientific literature, environmental benchmarks, and mission-specific requirements simultaneously. Satellite Data Analysis Systems powered by Retrieval-Augmented Generation (RAG) address this by d

Introduction Loan underwriting requires the rapid processing of vast financial documents, regulatory guidelines, and market data under tight deadlines, a challenge that rigid scoring models and manual review workflows are ill-equipped to handle. Underwriters must assess creditworthiness, collateral quality, and compliance requirements while keeping pace with constantly shifting lending regulations and economic conditions. Loan Underwriting Systems powered by Retrieval-Augment

Introduction Veterinary professionals must deliver accurate diagnoses across many species with unique biological differences, while keeping up with constantly evolving research and treatment guidelines. Retrieval Augmented Generation powered diagnostic systems provide real time access to veterinary literature, species specific protocols, diagnostic data, and patient history. By retrieving and synthesizing the most relevant and up to date evidence, these systems deliver contex

The High-Stakes Gamble: Why "Good Enough" Compliance is No Longer Enough In the modern global economy, data and digital operations are the engines of growth. But for the legal and risk teams tasked with managing them, these assets are like enriched uranium : immensely powerful when harnessed correctly, but catastrophic if mishandled. We have moved past the era where compliance was a back-office formality; today, it is the frontline of corporate survival. The Problem: A Labyri

Introduction Traditional AI interactions require rigid command structures limiting natural communication. Developers struggle to extract optimal responses from language models without specialized knowledge. Manual experimentation with different prompting approaches consumes significant development time. Inconsistent model outputs complicate production deployment and user experience. Llama 3:8B Chat transforms AI interactions through instruction-tuned conversational capabiliti

Introduction In this comprehensive tutorial, we'll build a binary image classification system to detect fungal infections in microscopy images of vine wood. We'll use Vision Transformers (ViT), a state-of-the-art deep learning architecture that applies transformer concepts to image classification. Dataset Overview Dataset: "An Eye on the Vine" This dataset comes from research on pathogen segmentation in vinewood fluorescence microscopy images. The dataset is available at:...

📋 Project Overview Domain : Real Estate Market Intelligence & Business Analytics Difficulty Level : Intermediate to Advanced Dataset Size : 100,000+ property transactions Timeline : 10-12 days Tools : Python, SQL, Power BI/Tableau Project Objective Design and implement a comprehensive real estate analytics solution that processes raw property data through a complete ETL pipeline, stores it in a structured database, and delivers actionable insights through interactive dashboa

Introduction The exponential growth of scientific literature, with millions of papers published annually, has made it increasingly difficult for researchers to keep pace with complex technical content. Traditional approaches based on manual reading and note taking create bottlenecks in knowledge discovery as scientists spend countless hours deciphering dense methodologies and synthesizing findings. Scientific Text Comprehension powered by Retrieval Augmented Generation (RAG)

Project Goal To implement machine learning within Elasticsearch for advanced anomaly detection and root cause analysis, enhancing system...

Introduction Project management systems handle vast numbers of tasks daily across organizations, creating complex workflows that demand...