ProjectMatch (projectmatch)

SIWES & Final Year Computer Science Project Finder, Scoping & Proposal Guidance CLI

CRITICAL DESIGN PRINCIPLE & INTENDED USE STATEMENT

IMPORTANT NOTICE: ProjectMatch RECOMMENDS and GUIDES — it NEVER generates or builds projects for you. It must never output actual project source code, scaffolded repositories, or complete implementations. Its job ends at: suggesting project ideas, describing scope, suggesting tech stacks, breaking down milestones, helping you articulate supervisor approval reasoning, and preparing you to explain and defend your own work. ProjectMatch does NOT write your actual proposal document or guarantee supervisor approval. You are expected to design, build, and be able to explain every line of code in your final submission.

Overview

Choosing a SIWES (Students Industrial Work Experience Scheme) or Final Year Project (FYP) under time pressure is daunting. Many Computer Science students either choose over-ambitious enterprise projects or overly trivial scripts that fail academic supervisor or defense panel scrutiny.

ProjectMatch is a zero-configuration, zero-cost, offline-first CLI dev tool that matches undergraduate CS students with realistic, portfolio-worthy, and defensible project ideas based on project type, skill level, interests, and timeframe.

Key Features

• Curated Project Bank: 230+ structured, realistic CS projects covering Web, Mobile, Backend, Data Science, Systems/Networking, and Security.
• Two Distinct Project Types: Built-in support for both SIWES (160+ practical entries) and Final Year Projects (70+ academic entries with supervisor-approval reasoning).
• Rebalanced FYP Difficulty Curve: FYP entries reflect 400–500 level academic expectations: 16.7% Beginner, 41.7% Intermediate, and 41.7% Advanced, featuring empirical benchmarking, non-trivial engineering, and trade-off evaluations.
• Conceptual Understanding Prompts: Every project includes 3-5 curated prompts to give an AI assistant to master underlying concepts without generating project code.
• Per-Topic Supervisor Defense Reasoning: FYP matches include problem statements, suitability rationale, supervisor approval factors, and defense talking points.
• Proposal & Defense Guidance Commands: Dedicated static guidance commands (projectmatch proposal and projectmatch checklist) for proposal writing and panel interview preparation.
• Zero Setup & Zero Cost: Works 100% offline out-of-the-box with no API key or database required.
• Optional AI Layer: Wire up Anthropic, OpenAI, Gemini, or local Ollama for tailored AI recommendations beyond the curated bank (--ai flag).

Supported Project Types: SIWES vs Final Year Project (FYP)

ProjectMatch explicitly distinguishes between two major academic scoping milestones:

1. SIWES Industrial Training (siwes):

   • Demonstrates practical engineering skills learned during industrial placement.
   • Shorter completion timeframe (4-8 weeks).
   • Centered around an end-of-placement panel defense presentation.

2. Final Year Project (final-year):

   • Substantial academic thesis project requiring upfront supervisor approval via a written proposal.
   • Demands a clear problem statement, technical depth, and empirical or trade-off evaluation.
   • Rebalanced difficulty curve (16.7% Beginner, 41.7% Intermediate, 41.7% Advanced).
   • Displays four per-topic supervisor approval & defense fields during find:
     • Problem Statement: One clear sentence defining the specific problem or inefficiency addressed.
     • Why This Suits You: Explanation of why the topic matches your skill level and available timeframe.
     • Why a Supervisor Might Approve This: Convincing academic and feasibility arguments (phrased as "likely to be viewed favorably because..." — never a guarantee).
     • Defending Your Choice (If Asked): 2-3 concrete talking points to explain to your supervisor why you selected this topic.

Conceptual Understanding Prompts vs Code Generation

Every project in ProjectMatch includes a dedicated Understanding Prompts section. These prompts are designed for students to ask an AI assistant (like ChatGPT, Claude, or local Ollama) to master concepts required by the project.

CRITICAL WARNING: Using AI to generate your actual project implementation code undermines the entire point of building and defending your project. Prompts must teach concepts generically, never generate project source code.

Correct vs. Incorrect AI Prompting Patterns

| Usage Type | Prompt Example | Why | | :--- | :--- | :--- | | CORRECT (Conceptual Learning) | "Explain how JWT authentication works and why refresh tokens matter for security" | Teaches fundamental security concepts without generating project-specific code. | | CORRECT (Generic Tradeoffs) | "What are the tradeoffs between relational and document databases using a generic example?" | Builds theoretical knowledge needed to defend architecture choices to supervisors. | | INCORRECT (Code Generation) | "Write a function that validates a JWT token in Express" | prohibited: Generates copy-paste code that invalidates student authorship. | | INCORRECT (Schema Generation) | "Generate the database schema for my library project" | prohibited: Bypasses the core design effort required for project defense. |

Using ProjectMatch at Scale (For Large Cohorts & Supervisors)

ProjectMatch works 100% offline out of the box using its curated 230+ project bank with score-based subset sampling.

For academic supervisors coordinating large student cohorts (100+ students per year) who want maximum project variety across their group:

• --ai Flag for Unlimited Variety: Passing --ai when running projectmatch find --ai queries an AI provider to synthesize additional tailored project concepts beyond the fixed bank.
• Free Local Setup with Ollama: Cost is never a barrier. You can configure a local, completely free LLM via Ollama (no API key or subscription required):

  projectmatch config set-provider ollama
  projectmatch config set-model llama3
  projectmatch config set-url http://localhost:11434

Installation

# Global installation via npm
npm install -g projectmatch

# Or run directly via npx without installation
npx projectmatch --help

Command Reference

1. projectmatch find [--ai]

Launch the interactive project matching wizard:

• Prompts for project type (siwes or final-year), skill level, interest area, available timeframe (weeks), and known languages/tools.
• Filters and ranks the project bank to deliver top recommendations.
• Displays milestones, conceptual Understanding Prompts, and supervisor approval fields for FYP.
• Includes --ai flag to fetch additional AI-generated recommendations if an LLM provider is configured.

projectmatch find

2. projectmatch list [--type <type>] [--category <cat>] [--level <level>]

Browse the curated project bank directly with optional filters:

• --type, -t: siwes, final-year
• --category, -c: web, mobile, backend, data, systems, security
• --level, -l: beginner, intermediate, advanced

# List all Advanced Final Year Projects in Backend Development
projectmatch list --type final-year --category backend --level advanced

3. projectmatch proposal

Display general guidance for writing a Final Year Project proposal for supervisor approval.

projectmatch proposal

4. projectmatch checklist

Display the SIWES final-defence readiness checklist and self-evaluation panel questions.

projectmatch checklist

5. projectmatch config <action> [value]

Configure optional AI provider settings (stored locally at ~/.projectmatch/config.json).

# Show current config
projectmatch config show

# Set provider to local free Ollama
projectmatch config set-provider ollama
projectmatch config set-model llama3
projectmatch config set-url http://localhost:11434

License

MIT (c) 2026 Fayol Smith