Startups often operate in high uncertainty environments and non-technical founders often find themselves in a position to make critical decisions for technical requirements. However, non-technical founders lack the ability to articulate these decisions which often leads to failure in the early stages of their startup journeys. Generative Ai offers a potential solution to non-technical founders yet current tools and methods fail them by being too generic or confident. This thesis dives into how an Ai-powered concept should be designed to bridge the gap between intention and technical requirements. The research includes a mixed methods approach and combines a literature review on start-up methodologies and frameworks with primary data collection. The survey revealed that 45% of participants identified communication with tech teams and technical jargon as a barrier for them. Semi-structured interviews further validated four recurring needs; the translation of intention into technical components, realistic estimation of cost, time and effort, aid in MVP scoping to prevent optimism bias and transparency mechanics to facilitate trust. As an answer to these findings, the thesis proposes a technical scoping assistant concept acting as a semantic translator and constraint-aware planner. The proposed concept includes an adversarial interaction to enforce prioritization and draw out missing information, and a “Glass Box” transparency layer to present assumptions, evidences and confidence level. The thesis contributes design criteria for Ai tools in startup context and provides a concept for a transparent and execution oriented assistant instead of ideation.
Le startup operano spesso in ambienti ad alta incertezza e i fondatori non tecnici si trovano frequentemente a dover prendere decisioni critiche in merito ai requisiti tecnici. Tuttavia, i fondatori non tecnici spesso non possiedono le competenze necessarie per articolare tali decisioni, il che porta frequentemente al fallimento nelle fasi iniziali del loro percorso imprenditoriale. L'Intelligenza Artificiale (IA) generativa offre una potenziale soluzione, eppure gli strumenti e i metodi attuali risultano inadeguati in quanto troppo generici o eccessivamente sicuri delle proprie risposte. Questa tesi analizza come dovrebbe essere progettato un sistema basato sull'IA per colmare il divario tra le intenzioni imprenditoriali e i requisiti tecnici. La ricerca adotta un approccio a metodi misti (mixed-methods) e combina una revisione della letteratura sulle metodologie e i framework per startup con una raccolta di dati primari. Il sondaggio ha rivelato che il 45% dei partecipanti identifica la comunicazione con i team tecnici e il gergo tecnico come un ostacolo principale. Le interviste semi-strutturate hanno ulteriormente validato quattro necessità ricorrenti: la traduzione delle intenzioni in componenti tecnici; la stima realistica di costi, tempi e sforzi; il supporto nella definizione del perimetro dell'MVP (scoping) per prevenire il pregiudizio di ottimismo (optimism bias); e meccanismi di trasparenza per favorire la fiducia. In risposta a questi risultati, la tesi propone il concept di un assistente per lo scoping tecnico che agisce come traduttore semantico e pianificatore consapevole dei vincoli (constraint-aware). Il concept proposto include un'interazione di tipo antagonistico (adversarial interaction) per forzare la prioritizzazione e far emergere le informazioni mancanti, nonché un livello di trasparenza "Glass Box" per presentare assunzioni, prove e livelli di confidenza. La tesi fornisce criteri di progettazione per gli strumenti di IA nel contesto delle startup e propone un concept per un assistente trasparente e orientato all'esecuzione, piuttosto che alla semplice ideazione.
Empowering non-technical founders: bridging technical gap with AI
Er, Gorkem
2025/2026
Abstract
Startups often operate in high uncertainty environments and non-technical founders often find themselves in a position to make critical decisions for technical requirements. However, non-technical founders lack the ability to articulate these decisions which often leads to failure in the early stages of their startup journeys. Generative Ai offers a potential solution to non-technical founders yet current tools and methods fail them by being too generic or confident. This thesis dives into how an Ai-powered concept should be designed to bridge the gap between intention and technical requirements. The research includes a mixed methods approach and combines a literature review on start-up methodologies and frameworks with primary data collection. The survey revealed that 45% of participants identified communication with tech teams and technical jargon as a barrier for them. Semi-structured interviews further validated four recurring needs; the translation of intention into technical components, realistic estimation of cost, time and effort, aid in MVP scoping to prevent optimism bias and transparency mechanics to facilitate trust. As an answer to these findings, the thesis proposes a technical scoping assistant concept acting as a semantic translator and constraint-aware planner. The proposed concept includes an adversarial interaction to enforce prioritization and draw out missing information, and a “Glass Box” transparency layer to present assumptions, evidences and confidence level. The thesis contributes design criteria for Ai tools in startup context and provides a concept for a transparent and execution oriented assistant instead of ideation.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis-Booklet-Gorkem-Er.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
95.56 MB
Formato
Adobe PDF
|
95.56 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/251393