IV-I is the lightest theory load of the whole CSE program on paper, but it’s where the curriculum hands over most of the steering wheel to the student — two Professional Electives with five or six subject choices each, two Open Elective slots pulled from outside the department, a hands-on Prompt Engineering skill course, and the start of the industry internship/mini-project track that runs into IV-II. Below is the official R23 course structure table followed by a subject-by-subject breakdown, including every elective option JNTUK lists for this semester.
Course Structure
| # | Category | Subject | L-T-P | Credits |
|---|---|---|---|---|
| 1 | Professional Core | Deep Learning | 2-1-0 | 3 |
| 2 | Management Course-II | Human Resources & Project Management | 2-0-0 | 2 |
| 3 | Professional Elective-IV | Software Architecture & Design Patterns / Blockchain Technology / Augmented Reality & Virtual Reality / Internet of Things / Agentic AI / 12-week SWAYAM-NPTEL MOOC (BoS-recommended) | 3-0-0 | 3 |
| 4 | Professional Elective-V | Agile Methodologies / Generative AI / Computer Vision / Cyber Physical Systems / 12-week SWAYAM-NPTEL MOOC (BoS-recommended) | 3-0-0 | 3 |
| 5 | Open Elective-III | Filled from the university-wide open elective pool | 3-0-0 | 3 |
| 6 | Open Elective-IV | Filled from the university-wide open elective pool | 3-0-0 | 3 |
| 7 | Skill Enhancement Course | Prompt Engineering / SWAYAM Plus — Certificate Program in Prompt Engineering and ChatGPT | 0-1-2 | 2 |
| 8 | Audit Course | Constitution of India | 2-0-0 | Non-credit |
| 9 | Internship | Evaluation of Industry Internship / Mini Project | – | 2 |
| Total | 18-2-2 | 21 |
Note: the same table also carries three optional add-on tracks students can pick up alongside the core 21 credits — a Minor Course (3-0-0-3, drawn from the same specialization-minor pool) and two Honors Course slots (3-0-0-3 each, drawn from the same honors pool). These aren’t part of the mandatory 21-credit total; they’re for students separately pursuing a Minor or Honors distinction in CSE.
On Open Elective III & IV: the R23 CSE document doesn’t lock these two slots to specific CSE subjects — they’re filled from whatever open elective pool the university offers that semester, typically courses run by other engineering departments. For reference, the same document separately lists what CSE itself teaches to other branches as open electives: Principles of Operating Systems/Computer Organization & Architecture, Principles of Database Management Systems, Object Oriented Programming through Java, and Principles of Software Engineering/Computer Networks — but that’s CSE’s outbound offering, not the inbound pool CSE students draw IV-I’s OE-III/OE-IV from.
Subject-Wise Syllabus
Deep Learning
— the theory backbone behind almost every neural-network question in a placement interview, tracing the field from classical ML up through CNNs, RNNs, and generative models.
- Unit 1: The roots of machine learning before deep learning took over — probabilistic modeling, early neural nets, kernel methods, decision trees, random forests, and gradient boosting — plus the overfitting/underfitting tradeoff that motivates going deeper.
- Unit 2: Biological vision and language versus their machine counterparts, the anatomy of artificial neural networks, and the mechanics of training and improving deep networks.
- Unit 3: Building and training networks with Keras, TensorFlow, Theano, and CNTK, worked through two classic examples — binary sentiment classification on movie reviews and multiclass classification on newswire topics.
- Unit 4: Convolutional neural networks (representation learning, convolutional layers, multichannel convolutions) and recurrent neural networks, implemented hands-on in PyTorch.
- Unit 5: Applied deep learning across machine vision, NLP, and generative adversarial networks, plus deep reinforcement learning and a look at generative research — autoencoders, Boltzmann machines, and deep belief networks.
Human Resources & Project Management
— the one paper in the semester that has nothing to do with writing code, covering how organizations hire and develop people, and how projects actually get delivered.
- Unit 1: HRM fundamentals — scope, functions, and emerging trends like e-HRM and HR audits — alongside workforce planning, job design, recruitment, and selection procedures.
- Unit 2: HR development and training methods, performance appraisal techniques, career counseling, and team/group dynamics.
- Unit 3: Core project management concepts — resource management, project types, project networks, life cycle stages, and how projects get appraised and selected.
- Unit 4: Matching management strategy to project type, plus the practical side of implementation — organizational forms, planning, control, and the human factors in delivery.
- Unit 5: Closing out a project — implementation prerequisites, performance review, and abandonment analysis for deciding when a project should be pulled.
Professional Elective-IV (choose one)
Five named subjects plus a BoS-recommended 12-week MOOC option — students pick one for 3 credits.
Software Architecture & Design Patterns
— moves students from writing classes to justifying why the classes are shaped that way, built around the classic Gang-of-Four pattern catalog and UML modeling.
- Unit 1: What a design pattern is, how the standard pattern catalog is organized, and the core ideas behind object-oriented analysis and design.
- Unit 2: Systems analysis in practice — gathering functional requirements, defining conceptual classes and relationships, and translating domain knowledge into a design.
- Unit 3: Structural design patterns — Adapter, Bridge, Composite, Decorator, Facade, Flyweight, and Proxy.
- Unit 4: The Model-View-Controller pattern applied end-to-end through a drawing-program case study, including undo handling and adding new features without breaking the design.
- Unit 5: Distributed object systems — client-server architecture, Java RMI, SOAP/REST web services, and the Enterprise Service Bus.
Blockchain Technology
— for students who want to understand what’s actually inside a block before they write a smart contract, covering consensus mechanisms, chain types, and security.
- Unit 1: Blockchain fundamentals and components, consensus protocols, blockchain types, and cryptocurrency basics (Bitcoin, altcoins, tokens).
- Unit 2: Public blockchain systems — Bitcoin and Ethereum — plus smart contracts and the oracle systems that feed them external data.
- Unit 3: Private and consortium blockchains, permissioned-chain algorithms, Hyperledger/Ripple/Corda, and how Initial Coin Offerings work.
- Unit 4: Blockchain security and privacy challenges, identity management, regulatory compliance, and applications across banking, healthcare, energy, and supply chain.
- Unit 5: Case studies in retail, banking, healthcare, and energy, plus hands-on blockchain development using Python and Hyperledger Fabric.
Augmented Reality & Virtual Reality
— a systems-level tour of AR/VR that moves from tracking hardware and display optics into the human perception science that makes immersion convincing.
- Unit 1: AR foundations — definitions, history, and displays — plus the tracking, calibration, and registration pipeline that anchors virtual content to the real world.
- Unit 2: Computer vision techniques for AR tracking (marker-based and natural-feature), interaction methods, and the software architectures behind AR applications.
- Unit 3: VR foundations — history, human physiology, the geometry of virtual worlds, and the optics of light, lenses, and displays.
- Unit 4: Human visual physiology from the eye to the visual cortex, depth/motion/color perception, and rendering techniques including ray tracing and distortion correction.
- Unit 5: Motion and the vestibular system, locomotion and social interaction in virtual worlds, and how spatial audio is modeled and rendered.
Internet of Things
— ties sensors, connectivity protocols, and edge processing together into one end-to-end IoT stack.
- Unit 1: IoT’s predecessors (wireless sensor networks, machine-to-machine communication), how modern IoT emerged, and its networking and addressing components.
- Unit 2: Sensing and actuation fundamentals, sensor/actuator characteristics, and IoT data-processing topologies and offloading strategies.
- Unit 3: Connectivity protocols spanning short and long range — Zigbee, Thread, RFID, NFC, LoRa, Wi-Fi, Bluetooth, Sigfox — and the communication/discovery/data protocol stack.
- Unit 4: IoT interoperability standards and frameworks, plus fog computing architecture and its applications.
- Unit 5: Emerging IoT paradigms and open challenges, illustrated with agricultural and vehicular IoT case studies.
Agentic AI
— the newest option on the list, covering how AI systems move from answering single prompts to autonomously planning and executing multi-step goals.
- Unit 1: What makes AI “agentic” versus reactive or purely generative — agent architectures, rationality, and the motivation for autonomous systems.
- Unit 2: Decision-making and planning under uncertainty, reinforcement learning fundamentals (MDPs, policy, reward shaping), and deep RL methods like PPO, DQN, and A3C.
- Unit 3: How large language models power agent reasoning — transformer basics, prompting strategies (zero-shot, chain-of-thought, ReAct), retrieval-augmented generation, and tool invocation.
- Unit 4: Practical agent frameworks — LangChain, LangGraph, AutoGen, CrewAI — plus vector databases, multi-agent coordination, and deployment/monitoring concerns.
- Unit 5: Responsible AI design for autonomous agents — bias, fairness, transparency, and safety/alignment — with case studies in healthcare, cybersecurity, and business automation.
Professional Elective-V (choose one)
Four named subjects plus a BoS-recommended 12-week MOOC option — students pick one for 3 credits.
Agile Methodologies
— less about a specific tool and more about the mindset shift from plan-driven delivery to iterative, feedback-driven software development.
- Unit 1: The Agile Manifesto and the values behind it, and why “agile” resists being reduced to one fixed process.
- Unit 2: The 12 Agile principles applied to real project delivery, team communication, and continuous improvement.
- Unit 3: Scrum — roles, ceremonies, sprint planning and retrospectives, and the self-organizing team model.
- Unit 4: Extreme Programming (XP) — its core practices and values, and how it embraces changing requirements through simple, incremental design.
- Unit 5: Lean thinking and Kanban — eliminating waste, managing flow, and the role of an Agile coach in driving organizational change.
Generative AI
— a survey of the models behind the current AI boom, from language models to image and music generation, including the frameworks used to build with them.
- Unit 1: Generative AI’s history and how it differs from discriminative modeling — GANs, VAEs, autoregressive models, and diffusion models — plus ethics and responsible-AI considerations.
- Unit 2: Language models and transformer architecture, text generation with models like BERT and GPT, prompt engineering, RLHF, and retrieval-augmented generation.
- Unit 3: Image generation — GANs, variational autoencoders, stable diffusion, and transformer-based image models like DALL-E and GPT-4V.
- Unit 4: Generative models for painting, music, and gameplay — style transfer, CycleGAN, music-generating RNNs, and RL-based agents.
- Unit 5: Fine-tuning and deploying open-source generative models — LangChain, Llama, Hugging Face, and transfer-learning workflows.
Computer Vision
— the math-heavy elective on this list, covering cameras, optics, and the geometry that turns 2D pixels into 3D understanding.
- Unit 1: Camera models, radiometry, shading, and recovering surface color from image color.
- Unit 2: Linear filters and convolution, Fourier analysis, edge detection, and texture representation/synthesis.
- Unit 3: Multi-view geometry, stereopsis, and image segmentation by clustering.
- Unit 4: Model-fitting techniques — the Hough transform, robust fitting, EM-based segmentation, and object tracking with Kalman filters.
- Unit 5: Geometric camera calibration and model-based vision, with case studies in mobile robot localization and medical image registration.
Cyber Physical Systems
— for students headed toward embedded or industrial systems, covering how software controllers stay correct, secure, and synchronized when tied to physical processes.
- Unit 1: Symbolic synthesis techniques for building verified controllers for cyber-physical systems.
- Unit 2: Security threats and countermeasures specific to cyber-physical systems, including system-theoretic defense approaches.
- Unit 3: Synchronization challenges in distributed CPS — consensus algorithms, time-triggered architectures, and lockstep execution.
- Unit 4: Real-time scheduling under fixed timing constraints, memory effects, and multicore scheduling under variability.
- Unit 5: Model integration across CPS design languages — causality, timing semantics, and formal language integration.
Skill Enhancement Course: Prompt Engineering
— a hands-on course built around getting reliable, structured output from LLMs, with a lab session paired to nearly every unit of theory.
- Unit 1: Prompt anatomy and the iterative prompting lifecycle, practiced through baseline-vs-enhanced prompt comparisons and diagnosing common failure modes.
- Unit 2: Advanced prompting patterns — few-shot examples, role-based personas, negative prompting, and constraint enforcement — tested through zero-shot vs. few-shot lab comparisons.
- Unit 3: Structured outputs and reasoning — generating valid JSON/YAML, chain-of-thought prompting, and task decomposition — validated with parser-checked exercises.
- Unit 4: Retrieval-augmented generation — RAG architecture, LangChain/LCEL basics, embeddings and vector stores — built hands-on into a working retrieval pipeline.
- Unit 5: LLM agents, multimodal prompting, and evaluation — tool-calling agents, text-to-image workflows, LLM-as-judge scoring, and prompt-injection safety testing.
Lab work runs alongside every unit rather than as a separate block: students install and configure LLM APIs, iterate on prompts across multiple rounds, build a JSON/YAML-validated output pipeline, assemble a working LCEL-based RAG chain, and finish by wiring up a tool-calling agent and running a prompt-injection safety test.
Audit Course: Constitution of India
— a mandatory, non-credit civics course every JNTUK student takes before graduating, covering how the Indian Constitution was built and how it functions today.
- Unit 1: How the Constitution was drafted, and its foundational philosophy — the Preamble and its salient features.
- Unit 2: Fundamental rights and duties — equality, freedom, protection from exploitation, religious freedom, constitutional remedies — alongside the Directive Principles of State Policy.
- Unit 3: The three organs of government — Parliament, the Executive (President, Governor, Council of Ministers), and the Judiciary — their composition and powers.
- Unit 4: Local self-government — district administration, municipalities, and the Panchayati Raj system across zila, block, and village levels.
- Unit 5: The Election Commission’s role and functioning, and the constitutional bodies safeguarding SC/ST/OBC and women’s welfare.
Internship Evaluation of Industry Internship / Mini Project
— not a taught subject but a 2-credit evaluation slot that grades the industry internship or mini-project every student is expected to complete this year, assessed against a rubric rather than classroom units.
Open Elective III & IV
Both are 3-0-0-3 slots pulled from the broader university open elective pool rather than a fixed CSE syllabus, so the exact subject a student sits for depends on what’s on offer that semester from other departments. Students should confirm the live options with their department before registration.