Prototyping
Prototyping is the practice of creating and evaluating a preliminary or partial implementation to explore or validate an idea, design, or approach.
Designing Deep Learning Systems: A guide for software engineers
Prototyping is an experimental phase for evaluating candidate algorithms and building proof-of-concept models against a business requirement.
- How it is used
- Bridges candidate algorithms and product requirements by iterating toward a deployable deep learning feature.
- How deep it goes
- Details an exploratory loop of data selection, experimental training, evaluation, stabilization, and handoff; defers production workflow construction.
- Examples
- Uses a security-camera motion-detection improvement scenario and visual-recognition experiments with ImageNet in JupyterLab.
- What you can do after
- Apply it via notebook-based experimentation to stabilize training and inference code for packaging and productionization.
Read how Designing Deep Learning Systems: A guide for software engineers explains prototyping →
Effective Platform Engineering
A prototype is an early interactive or visual representation of a product feature used to explore and evaluate a design.
- How it is used
- Tests proposed platform experiences and technology choices before production commitments.
- How deep it goes
- Outlines prototype tools, feedback deliverables, and realistic proof-of-concept criteria without specifying a complete implementation procedure.
- Examples
- Uses an interactive critical-feature prototype, a Figma mockup, and a proof of concept on actual platform infrastructure.
- What you can do after
- Practice/perform prototype-based validation, gathering user feedback or testing a technology before committing to full implementation.
Read how Effective Platform Engineering explains prototyping →
Design for Developers
A prototype is a preliminary version of a product or feature created for evaluation before implementation.
- How it is used
- Tests proposed website or application experiences with users before development to clarify functionality and usability.
- How deep it goes
- Distinguishes low- and high-fidelity prototypes and explains their testing, interaction, and iteration roles without implementation detail.
- Examples
- Uses pen-and-paper sketches, low-fidelity wireframes, clickable wireframes, refined mock-ups, and guided user walkthroughs.
- What you can do after
- Apply it to test concepts early, revise user flows from feedback, and reduce costly late redesigns.
Everyday Data Visualization
A prototype is an early-stage version of a design created to test concepts or demonstrate potential options rather than serve in production.
- How it is used
- Uses prototypes to test dashboard concepts, demonstrate alternatives, and collect stakeholder or user feedback before production.
- How deep it goes
- Explains a practical workflow from representative data through rough implementation, interaction testing, and feedback, without prescribing a specific visualization tool.
- Examples
- Uses rough dashboard builds, survey-like sample data, and a stakeholder-facing prototype built with real data.
- What you can do after
- Apply it via a chosen visualization tool to test interactions and demonstrate a dashboard design quickly.
The Art of AI Product Development
A prototype is an early or simplified version of a product or feature that demonstrates its behavior for learning and evaluation.
- How it is used
- Uses early working artifacts to test feasibility, align product requirements, and learn from users before committing to full delivery.
- How deep it goes
- Details rapid and UX-centered iteration, including realistic outputs and simulated failures, but stops short of implementation specifics.
- Examples
- Uses personalized playlists, clickable interfaces, realistic generated reports, and simulated hallucinations, regulatory errors, uncertain data, and formatting defects.
- What you can do after
- Evaluate AI opportunities and user responses, then iterate models and interfaces using evidence about usefulness, trust, errors, and adoption.
Read how The Art of AI Product Development explains prototyping →
Lead Developer Career Guide
The practice of creating an early, simplified model of a product or feature to evaluate feasibility and gather feedback.
- How it is used
- Uses early product models to test ideas, gather feedback, learn technologies, and communicate with teams and stakeholders.
- How deep it goes
- Defines prototypes and explains their forms, benefits, and reuse, without prescribing a formal evaluation procedure.
- Examples
- Uses paper sketches, mockups, wireframes, coded projects, CRUD applications, and a Hello World application as examples.
- What you can do after
- Apply prototyping to test ideas early, collect feedback, explore technologies, and communicate a project’s value.
Delivering Machine Learning Projects: From design to deployment
A prototype is a preliminary implementation created to demonstrate and evaluate a proposed system or interaction.
- How it is used
- Positions a visible concept demo alongside system building to demonstrate customer value before stakeholders cancel or redirect delivery.
- How deep it goes
- Develops concurrent UX construction, user testing, and review preparation, but stops short of specifying production implementation details.
- Examples
- Uses frontend UX components, user tests, a stakeholder review session, and a concept demo combining application and model work.
- What you can do after
- Evaluate a proposed system’s customer value through a reviewable interaction before committing to full production delivery.
Read how Delivering Machine Learning Projects: From design to deployment explains prototyping →
Effective Data Analysis
Rapid prototyping is an approach that creates and iteratively refines preliminary versions of a product or deliverable.
- How it is used
- Uses data-powered apps to develop and refine custom analytical deliverables through early stakeholder feedback.
- How deep it goes
- Defines preliminary versions, iteration, and feedback, but does not detail a complete prototyping process or evaluation framework.
- Examples
- Uses interactive data-powered apps, including a Streamlit rat-sightings prediction app, as concrete prototypes.
- What you can do after
- Apply it via data-powered apps to create preliminary deliverables, gather feedback, and refine interactive analytical products.
Related concepts
- requirementsproduces · A requirement is a stated need, capability, quality, condition, or constraint that a system, product, solution, or project is expected to satisfy.
- Streamlitused by · Streamlit is an open-source Python framework for building interactive web applications, particularly for data and machine-learning applications.
- agile architectureused by · Agile architecture is an iterative approach to architectural design that incorporates collaboration, experimentation, feedback, and emergent decisions.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.