Computer Science Learning Journey: Years 10 & 11.
An engineering curriculum designed for the modern compute paradigm. Students master rigorous computational theory, data structures, and algorithm complexity for Edexcel Paper 1 & 2, while building and deploying live AI applications using Python, Docker containers, vector databases, and cloud infrastructure.
Foundational Systems, Algorithmic Logic & AI Microservices
Core iGCSE theory, binary arithmetic, robust Python 3 data structures, and the inception of local AI APIs.
Data Representation: Binary, Hex & Digital Media
Master binary, two's complement signed integers, overflow errors, and hexadecimal conversion. Analyze how image pixels, color depth, metadata, and sound sample rates map to physical memory.
Core iGCSE Objectives:
- Convert seamlessly between Denary, 8-bit Binary, and Hexadecimal.
- Perform binary addition and identify arithmetic overflow flags.
- Calculate file sizes for bitmap images: \(\text{Height} \times \text{Width} \times \text{Colour Depth}\).
- Sound digitization: sample rate, sample resolution, and lossy vs. lossless compression.
pyproject.toml and Git.Algorithmic Thinking & Python 3 Onscreen Foundations
Decomposition, abstraction, and pseudocode standard formats. Write syntactically pristine Python 3
utilizing variables, arithmetic operators (//, %), selection (if/elif/else),
and count/condition-controlled loops.
Core iGCSE Objectives:
- Develop flowchart and standard pseudocode algorithmic solutions.
- Employ nested
whileandforloops with strict boundary condition checks. - Input validation: range checks, presence checks, and type checks.
- Implement trace tables to debug dry-run execution and find logic errors.
pytest verifying all edge cases.Local LLM Runtime & Prompt Engineering Pipelines
Demystify Large Language Models from the command line. Configure local inference engines (Ollama running Llama 3 / Mistral), study tokenization, temperature parameters, and build Python scripts interacting with local model endpoints.
Catalyst AI Engineering Objectives:
- Understand Transformer architectures: tokens, context windows, and embedding vectors.
- Configure local inference runtimes without external cloud API dependencies.
- Construct structured JSON outputs from LLMs using deterministic system prompts.
- Evaluate hallucinations, prompt injection vulnerabilities, and system boundaries.
Data Structures: 1D/2D Lists, Dictionaries & Strings
Deep dive into Python data structures: indexed collections, matrix navigation in 2D arrays, associative dictionaries, and complex string slicing/sanitization methods.
Core iGCSE Objectives:
- Manipulate 1D and 2D arrays: traverse, aggregate, and update multi-dimensional data.
- String manipulation: substring extraction, concatenation, character ASCII conversions.
- Reading and writing structured records to external flat text and CSV files.
- Implement data defensive parsing with
try...exceptblocks.
Searching, Sorting & Algorithmic Complexity
Inspect, trace, and implement classic search and sort algorithms. Compare time and space efficiency: Linear vs. Binary search, Bubble sort vs. Merge sort.
Core iGCSE Objectives:
- Linear Search vs. Binary Search (pre-condition of sorted lists and \(O(\log n)\) efficiency).
- Trace step-by-step passes of Bubble Sort with early-exit swap flags.
- Understand divide-and-conquer principles underpinning the Merge Sort algorithm.
- Calculate algorithm iterations and comparison counts across worst-case and best-case arrays.
FastAPI AI Microservices: Building Backend Endpoints
Transition from simple scripts to production backend engineering. Build high-performance asynchronous REST APIs using FastAPI and Pydantic schemas to expose AI models to client applications.
Catalyst AI Engineering Objectives:
- RESTful API design principles: HTTP methods (GET, POST, DELETE) and status codes (200, 400, 404, 500).
- Strong typing and schema validation using Pydantic data models.
- Asynchronous Python request handling (
async / await) for non-blocking inference. - Integrating LLM streaming responses via Server-Sent Events (SSE).
/api/summarize that ingests documents and returns verified AI key takeaways.CPU Architecture & The Von Neumann Model
The inner mechanisms of computation. Examine the Fetch-Decode-Execute cycle, CPU registers (PC, MAR, MDR, ACC), buses (Address, Data, Control), cache levels, and factors determining processor performance.
Core iGCSE Objectives:
- Trace data flow between RAM, MAR, MDR, and ALU during the F-D-E machine cycle.
- Evaluate the impact of clock speed (GHz), cache memory size, and multi-core architecture.
- Differentiate primary volatile memory (RAM), non-volatile ROM/BIOS, and secondary storage technologies.
- Solid-state (flash) vs. Magnetic vs. Optical storage: capacity, speed, durability, and cost.
System Software, Memory Management & Boolean Logic
Operating system responsibilities: multitasking, scheduler algorithms, paging, virtual memory, and device drivers. Combine this with circuit logic: truth tables and logic circuits using AND, OR, NOT, and XOR gates.
Core iGCSE Objectives:
- Construct truth tables and simplify logic statements for multi-gate circuits.
- Analyze virtual memory swapping and thrashing when physical RAM is exhausted.
- Evaluate OS utility software: disk defragmentation, firewalls, and encryption utilities.
- High-level vs. Low-level languages: Compilers vs. Interpreters vs. Assemblers.
Docker Containerization & Environment Isolation
Solve "it works on my machine" permanently. Master container concepts: Linux namespaces, cgroups, writing
multi-stage Dockerfiles, and bundling Python AI microservices into immutable, reproducible artifacts.
Catalyst DevOps Objectives:
- Understand difference between Virtual Machines (hypervisors) and Lightweight Containers (shared OS kernel).
- Write optimized, secure
Dockerfiledefinitions using lightweight Alpine and Debian slim base images. - Manage environment variables, build arguments, and file system volume mounts.
- Build, tag, and publish container images to container registries (Docker Hub / GitHub Container Registry).
-p 8000:8000 and health checks.Networks, Vector Embeddings, Cloud CI/CD & Exam Synthesis
Advanced protocols, cyber defense, RAG architecture, cloud container deployment, and Pearson 4CP0 Grade 9 exam mastery.
Network Topologies, Hardware & The 4-Layer TCP/IP Model
How the modern Internet functions. Explore LAN vs. WAN, Star vs. Mesh topologies, routers, switches, and the 4-layer TCP/IP stack (Application, Transport, Internet, Link) alongside DNS resolution and packet switching.
Core iGCSE Objectives:
- Explain packet switching: headers (IP source/dest, packet seq), payload, and checksum error checking.
- Map key protocols to TCP/IP layers: HTTP, HTTPS, FTP, SMTP, IMAP, TCP, UDP, IP.
- Evaluate Star vs. Mesh network topologies regarding fault tolerance, cost, and cabling overhead.
- Examine wireless networking: frequency bands (2.4GHz vs 5GHz), channels, encryption (WPA3).
Cybersecurity, Cryptography & Social-Ethical Impacts
Threat identification: SQL injection, brute force, phishing, DoS/DDoS, and spyware. Defense mechanisms: symmetric/asymmetric encryption, firewalls, and penetration testing. Ethical evaluation of autonomous systems and copyright.
Core iGCSE Objectives:
- Identify cybersecurity vulnerabilities and formulate mitigation policies.
- Explain public-key asymmetric cryptography (RSA / ECC) vs. symmetric shared-secret ciphers.
- Ethical, legal, and environmental impacts of computer science (UK Data Protection Act, Computer Misuse Act).
- Environmental costs of computation: server farm energy consumption, e-waste, rare-earth mineral mining.
Retrieval-Augmented Generation (RAG) & Vector Databases
Ground AI models in real factual knowledge. Build an end-to-end RAG system: chunking large technical documents, computing multi-dimensional mathematical embeddings, storing in ChromaDB/Pinecone, and executing cosine similarity queries.
Catalyst AI Engineering Objectives:
- Understand mathematical embeddings: vectors representing semantic meaning in 1536-dimensional space.
- Calculate Cosine Similarity: \(\cos(\theta) = \frac{\mathbf{A} \cdot \mathbf{B}}{\|\mathbf{A}\| \|\mathbf{B}\|}\).
- Implement document chunking strategies with overlap to maintain contextual continuity.
- Synthesize prompt injection defense when assembling retrieved document context for LLMs.
Cloud Deployment: Serverless, AWS & Container Orchestration
Deploy software to global infrastructure. Provision cloud virtual servers, configure reverse proxies (Nginx / Traefik), configure SSL/TLS certificates via Let's Encrypt, and deploy multi-container systems using Docker Compose.
Catalyst DevOps Objectives:
- Define multi-service architectures using
docker-compose.yml(FastAPI + Redis + ChromaDB). - Configure reverse proxy routing, SSL termination, and rate limiting headers.
- Deploy containerized workloads to cloud hosts (AWS Lightsail, Railway, or Google Cloud Run).
- Configure DNS A-records and CNAMEs mapping public domains to running services.
CI/CD Pipelines: GitHub Actions & Automated Testing
Automate the software release cycle. Create GitHub Actions workflows that automatically run linting (Ruff/Flake8), execute unit test suites on every git push, build Docker images, and deploy automatically upon passing tests.
Catalyst DevOps Objectives:
- Write YAML-based CI/CD workflows triggered by pull requests and main branch merges.
- Automate code quality: static analysis, type checking (
mypy), and security auditing. - Automate container registry push using encrypted repository secrets.
- Implement zero-downtime deployment rollouts and health check confirmations.
.github/workflows/deploy.yml pipeline that tests and deploys code in under 90 seconds.AI Observability, Latency Optimization & Safety Guardrails
Manage AI in production. Track latency metrics (Time To First Token), token consumption costs, model drift, and implement programmatic guardrails (NeMo Guardrails / regex filters) protecting against jailbreaks and privacy leaks.
Catalyst AI Engineering Objectives:
- Measure and optimize LLM inference metrics: TTFT (Time To First Token) and generation tokens/sec.
- Implement streaming token pipelines to enhance perceived user experience.
- Construct defensive guardrail filters evaluating user prompts before hitting core LLMs.
- Log and monitor API telemetry using structured logging and OpenTelemetry tracing.
Paper 2 Python 3 Onscreen Programming Mastery
Rigorous timed drills under official Pearson Edexcel onscreen coding examination conditions. Master rapid debugging, file handling, string sanitization, and 15-mark algorithmic design questions.
Core iGCSE Objectives:
- Deconstruct and debug pre-written flawed Python source files under strict time constraints.
- Implement robust text/CSV read and write routines with clean close handlers.
- Execute multi-stage validation routines: length, range, format, lookup, and presence checks.
- Deliver production-quality clean code conforming exactly to Pearson Edexcel marking schemas.
Paper 1 Theory Mastery & Extended Evaluative Writing
Final consolidation of computer science principles: binary calculations, logic gates, network topologies, and 6-mark/8-mark extended evaluative essays addressing legal, ethical, and societal implications of computing.
Core iGCSE Objectives:
- Flawless execution of binary two's complement and hexadecimal mathematical problems.
- Detailed technical explanations of TCP/IP protocol layers and packet-switched routing.
- High-scoring structural frameworks for 8-mark extended ethical and legal questions.
- Rapid recall of CPU architecture, primary vs. secondary storage, and OS memory management.
Senior AI Software Capstone: Live Production Defense
The pinnacle of the Catalyst Computer Science track. Defend an original, fully deployed, containerized AI software system before a panel of industry software engineers and faculty examiners.
Capstone Requirements:
- Complete Git repository with clear README, modular architecture, and automated test coverage.
- Publicly accessible HTTPS cloud URL running via containerized cloud infrastructure.
- Technical oral defense addressing architecture choices, Big-O complexity, and ethical boundaries.
- Integration of custom dataset, fine-tuned prompt embeddings, or specialized heuristic algorithms.
The Engineer's Lexicon & Architecture Vault
Essential theoretical constructs, complexity benchmarks, and deployment patterns every Catalyst Computer Science student commits to memory for Grade 9 exam mastery and software engineering excellence.
Method of representing signed negative integers in binary. Invert all bits (one's complement) and add 1.
Divide-and-conquer search on sorted arrays. Halves the search space every iteration.
Bubble sort repeatedly swaps adjacent elements. Merge sort recursively splits and combines sorted lists.
PC holds address of next instruction. Address moves to MAR via Address Bus. Instruction fetched into MDR via Data Bus.
Hierarchical model for network communication: Application (HTTP/DNS), Transport (TCP/UDP), Internet (IP/ICMP), Link (Ethernet/Wi-Fi MAC).
Calculates cosine of angle between two embedding vectors in multidimensional space to quantify semantic similarity.
Retrieval-Augmented Generation: query → generate query vector → retrieve top-K semantic chunks from DB → augment prompt → LLM inference.
Using separate build stages to compile dependencies, copying only lean production binaries into the final image to minimize attack surface.
Pearson Edexcel iGCSE (4CP0) & Catalyst Assessment Blueprint
Complete breakdown of formal examination papers and senior engineering requirements.
| Assessment Component | Format & Environment | Duration | Total Marks | Weighting | Core Competencies Assessed |
|---|---|---|---|---|---|
|
Paper 1: Principles of Computer Science Code: 4CP0/01 |
Written Examination (Pen & Paper) | 2 Hours | 80 Marks | 50% of iGCSE | Computational thinking, data representation (binary, hex, media), hardware architecture (CPU, memory, storage), networks, network security, software types, and 8-mark extended ethical/environmental impact essays. |
|
Paper 2: Application of Computational Thinking Code: 4CP0/02 |
Onscreen Practical Examination (Python 3 IDE) | 2 Hours | 80 Marks | 50% of iGCSE | Hands-on onscreen coding: modifying flawed starter code, writing algorithms from scratch, file input/output processing, data validation routines, 1D/2D arrays, and testing edge cases. |
|
Catalyst Senior AI Systems Capstone CIS Diploma Requirement |
Live Cloud Deployment & Oral Defense Panel | Continuous + 20 Min Defense | Graded (Distinction / Merit / Pass) | Catalyst Diploma | Original full-stack software application running live in the cloud via Docker, incorporating verified local/cloud LLM intelligence, automated CI/CD GitHub workflows, and technical oral defense. |