Define
Role, objective, task, behavior and the scope of the request.
You know your field. You know the problem you need to solve. You know which constraints matter, which sources can be trusted, which assumptions are acceptable and what kind of result you need.
But communicating all of that clearly to an AI is another matter.
A short or poorly structured prompt can turn professional knowledge into an incomplete request — leaving the AI to fill in gaps, make assumptions or interpret choices that should have remained yours.
GPS-AI helps prevent that.
It doesn't teach you prompt engineering, and it doesn't do your thinking for you. It helps you structure what you already know — defining the role, objective, task, constraints, sources, assumptions, input data, edge cases and expected output — and turns those choices into a structured instruction for AI. So the AI receives your expertise, not an approximation of it.
Desktop application + PDF Methodology Guide + reusable .gpsai projects
An incomplete prompt can leave the AI to fill gaps, introduce assumptions or interpret choices that should remain yours.
Role, objective, task, behavior and the scope of the request.
Constraints, sources, assumptions, variable data, edge cases and what must not be invented.
Response structure, file format and concrete output requirements.
GPS-AI helps turn professional requirements into a structured instruction while keeping expertise and responsibility in human hands.
They are not a prompt library to copy. They are a prompt library to understand.
GPS-AI breaks each example into its functional elements. You can see the role each part plays, modify it consciously and observe how it contributes to the final request.
The examples give you a starting point. The interface helps you understand their structure. The Methodology Guide explains why the method works. The expertise and decisions remain yours.
Personal criteria, current data, trade-offs and genuine ambiguity.
Source hierarchy, evidence quality, uncertainty and competing interpretations.
Explicit assumptions, costs, scenarios, current sources and multiple outputs.
Regulatory hierarchy, missing data, applicability and professional responsibility.
27 scenarios, 54 .gpsai projects — almost always paired as an everyday case and a technical/professional case, plus the four advanced examples (LIFE, SCIENCE, TECH, PROFESSIONAL) shown above. Open any item to see what type of prompt it is and who it's for.
Career-guidance consulting: identifies 3-5 development paths aligned with interests and skills in the AI era. For anyone choosing or reconsidering their career path.
Comparative analysis of three candidate cities for a medium-term relocation, covering cost of living, jobs, quality of life and bureaucracy. For anyone considering an international move.
Independent tech-purchase consulting: translates intended use into technical requirements and compares 2-3 real models. For anyone buying an expensive device without deep technical expertise.
Science-communication literature review: distinguishes established findings from preliminary or disputed ones. For anyone wanting an accurate picture of a complex scientific topic.
Point-by-point technical check of a solution against a sector standard, without replacing the licensed professional. For technicians and designers who need to verify compliance before proceeding.
Exploratory statistical analysis: finds patterns, anomalies and correlations, with a reusable Python script. For anyone with basic spreadsheet skills who wants a deeper analysis.
Comparative financial decision between renting and buying, making the medium-term impact explicit. For anyone deciding without specific financial training.
Balanced analysis of a scientific debate, with evidence and limits on both sides, favoring neither. For anyone wanting to understand a real debate without one-sided simplification.
Drafting a formal, well-documented complaint from scratch to improve the odds of a concrete response. For anyone disputing a service failure with a company or organization.
Formal documentation of a non-conformity found, traceable and compliant with required standards. For technicians and quality managers formalizing it for an external body.
Constraint-based planning: breaks the project into ordered phases, with timelines and overrun risk. For anyone tackling a renovation for the first time.
Technical planning of pre-operation checks, with purpose and acceptance criterion for each test. For technicians organizing commissioning before handover.
Everyday cause-effect diagnosis: lists the most likely causes and how to safely check them. For anyone wanting to understand a problem before calling a technician.
Industrial-grade cause-effect diagnosis: applies a structured method (e.g. 5 Whys) to trace the root cause. For process engineers and production managers.
Reorganizing disordered material: turns scattered notes into readable, standard-compliant minutes. For administrators or residents tasked with drafting them.
Turning raw data into a concise report highlighting trends, anomalies and threshold exceedances. For monitoring technicians and maintenance managers.
Adapting the communication register: explains simply and reassuringly how an online service works. For anyone helping an elderly or digitally inexperienced person.
Professional register adaptation: translates a technical/regulatory outcome into a clear client communication. For consultants and technicians explaining results to non-expert clients.
Preparing for a personal negotiation: supporting arguments, likely objections and responses. For anyone facing a negotiation over work, rent or a supply contract.
Preparing for a technical negotiation: builds a solid, documented position, anticipating likely objections. For technicians and consultants facing a review or certification body.
Everyday risk assessment: risks and opportunities of a significant personal choice, with probability and impact. For anyone deciding on a job change or new venture.
Industrial-grade risk assessment: safety, quality and operational-continuity risks before authorizing a change. For process engineers and production managers.
Constrained formal generation: turns a textual description into an accurate flowchart (phases, decisions, owners). For process analysts documenting an existing process visually.
Advanced multicriteria decision-support example: compares Lisbon, Berlin and Valencia on the same criteria. For anyone weighing a major life relocation, and to see GPS-AI applied to a complex case.
Advanced evidence-mapping example on a controversial topic (microplastics), spanning multiple study types. For scientifically literate readers wanting a rigorous, unsimplified analysis.
Advanced example with multiple output: compares three scenarios (local/cloud/hybrid) on performance, privacy and cost. Demonstrates the MULTI output option and multi-requirement prompts.
Advanced, maximum-complexity technical example: organizes legislation, grid connection and procedures for a 45 kWp PV system. For licensed electrical professionals handling real technical-regulatory cases.
Also included: a ready-to-use offline worksheet (01_EVERYDAY_Analisi_Bolletta_Elettrica_IT, .docx / .txt) — not a completed .gpsai project, but the marker-based form (see FAQ17) already filled in for a real case (reading an electricity bill), showing what a file ready to be re-imported into the app looks like.
Enabled by default, this section adds an instruction to the prompt telling the destination AI to treat pasted documents, sources and inputs as data to analyze, not instructions to execute — helping reduce the risk that external content alters the role, objective, constraints or information perimeter defined in the prompt.
It is a mitigation, not an absolute guarantee: no measure of this kind can eliminate the risk completely. But it adds a layer of protection that remains uncommon in tools designed for an individual professional.
The PDF Methodology Guide is not simply there to explain where to click. It explains why GPS-AI asks for specific information, what each choice does and how it can influence the AI's behavior.
Role, objective, sources, assumptions, response structure, output format, variable data and edge cases become deliberate choices rather than fields to fill in.
Keep the structure, requirements and data of your work in a project you can reopen.
The prompt doesn't remain disposable text: it becomes a reusable work asset that can be improved and adapted to recurring tasks or similar cases.
GPS-AI is not an artificial intelligence and does not include one. It builds the prompt text, which you then paste into the AI chat of your choice — ChatGPT, Claude, Gemini or any other text-based assistant.
No. GPS-AI doesn't require prompt-engineering expertise; it requires your expertise in your own field. The app guides you field by field toward information you already possess, not toward AI-writing techniques.
No. GPS-AI organizes and structures the information and choices you define into the final prompt. It does not replace your expertise by autonomously introducing objectives, constraints, sources or assumptions you did not establish.
Any text-based AI assistant — ChatGPT, Claude, Gemini and others. GPS-AI does not connect to any external service and makes no AI calls: it generates text that you copy and paste into the chat you choose.
No. GPS-AI prepares a structured prompt. You transfer it to the AI assistant you choose and evaluate the result yourself. This keeps the request-design tool separate from the AI model that executes it.
No. GPS-AI improves how precisely you communicate requirements, sources, constraints and data, but verification, validation and responsibility for the result remain with the author, especially in technical, regulatory or high-impact contexts.
Response structure defines how the content should be organized; output format defines the medium or file type required, such as PDF, DOCX, XLSX or Markdown. GPS-AI deliberately keeps these two choices separate.
Yes. The .gpsai project preserves the choices you made and can be reopened, reviewed, corrected and adapted, avoiding the need to rebuild a validated work structure from scratch.
No. It is a guide to understanding the choices. It explains why fields such as role, objective, constraints, sources, assumptions, edge cases and output requirements exist, and how they affect the quality and scope of the request sent to AI.
GPS-AI makes no calls to AI services; it generates the prompt text locally. You decide which content to paste into the AI assistant you use, and you remain responsible for applying the appropriate confidentiality and data-handling rules for that service.
No. The GPS-AI executable includes everything it needs to run: no additional component, such as the .NET Runtime, needs to be installed separately.
One-time purchase. GPS-AI is currently available at a launch price of €14.90 during its initial release period. You pay once and keep the file, with no renewals or subscriptions.
Yes, updates to the current version (1.x) are included at no extra cost. A possible future major version, with substantially new functionality, may be offered as a separate paid upgrade.
No. It is a mitigation, not an absolute guarantee. The function instructs the destination AI to treat pasted data as information to analyze rather than instructions to execute, reducing the risk that external content changes the AI's behavior — but no measure of this kind can eliminate the risk completely.
The GPS-AI Windows application, the PDF Methodology Guide integrated into the app, and the ability to generate, save and reopen an unlimited number of prompt projects (.gpsai files) together with the related text files ready to paste — with the option to export the completed prompt as a PDF for easier reading.
Yes. The interface is bilingual and you can switch languages at any time using the built-in selector.
Yes. The «TXT/DOCX Sheet» button works both ways: it creates a blank worksheet (TXT or DOCX) with a marked section for each project field, such as <OBJECTIVE>...</OBJECTIVE>, and it can also export a prompt you're already working on — with its fields already filled in — into that same marked format, so you can continue editing it offline, by hand or with someone else's help. Either way, importing the file back lets GPS-AI recognize the markers automatically and transfer the content into the corresponding project fields.
Your purchase includes an unlimited-time usage license, restricted to your own computers for your own personal or professional use. Distribution, transfer or sharing of the file, in any form, to third parties is not permitted.
To receive any updates you are entitled to under the purchased license, after buying you need to send an email to registrazione-gps-ai@lanzara.it. Registering allows us to send you communications about the updates you're entitled to, and does not change the terms of the license.
Desktop application for structured prompt design, with a PDF Methodology Guide and reusable projects.
Get GPS-AI