Jul 23 - Mar 24
Siemens' sales reps needed help gathering customer needs quickly, translating them into structured insights, and recommending tailored portfolio options in real time. We partnered with them to design the PortfolioXpert, an AI conversational tool that began as a digital companion and later evolved into an intelligent chatbot for conferences.
Full breakdown of the case study available on request
Copyright © Mattia Capozzi Sorrentino 2025
Copyright © Mattia Capozzi Sorrentino 2025
Copyright © Mattia Capozzi Sorrentino 2025
MattiaCapozziSorrentino
Copyright © Mattia Capozzi Sorrentino 2025
Copyright © Mattia Capozzi Sorrentino 2025
Copyright © Mattia Capozzi Sorrentino 2025
Context
Siemens PortfolioXpert was developed to simplify how Siemens sales representatives and customers interact with Siemens’ extensive industrial portfolio. Initially, it served as a structured tool to capture client needs, document conversations, and match solutions. Over time, the tool evolved into a conversational AI assistant, supporting real-time guidance during sales meetings and customer events.
Goal
The main goal was to make Siemens’ complex portfolio easier to navigate, both for sales representatives and for their clients. Conversations with clients were often inconsistent, time-consuming, and focused too much on reporting instead of value. We wanted to change that by creating a tool that could guide meetings in real time, surface the right information instantly, and build more trust between Siemens and their customers.
Impact
Team
I spent time understanding the world of Siemens sales reps and their clients: through 23 interviews, one survey, and two workshops (assumption-mapping workshop and AI ideation workshop), we uncovered a consistent set of challenges:
Conversations with clients were often unstructured.
The vast Siemens portfolio was difficult to navigate in real time.
Sales reps were losing hours on manual reporting instead of building relationships.
Product discovery
User Personas
As part of discovery, I created lightweight personas to keep our focus on the people using this tool every day.
For example, a senior sales representative juggling multiple client meetings in one week needed quick, structured support to stay consistent and credible in front of customers: mapping goals, frustrations, and feature needs helped the team align the team on what really mattered: efficient conversations, simplified portfolio navigation, and clear professional reports.
These personas became practical design anchors, guiding deliverables like journey maps, workshop boards, wireframes, and eventually the conversational prototypes that formed the backbone of the solution.
Ideation
With research insights and personas in place, I moved into ideation: the goal here was to make the assistant feel both conversational and structured, something that could guide a meeting without overwhelming it.
I sketched out early conversation flows, exploring how prompts, quick replies, and structured answers could work together. From there, I built low-fidelity wireframes that simulated real dialogues between sales reps and the AI.
These wireframes let us test different tones, interaction patterns, and layouts quickly: by putting these rough prototypes in front of stakeholders and reps, we could validate what felt natural in a live conversation and refine the flow before investing in high-fidelity design.
PortfolioXpert: a breakdown
The final product was a conversational AI assistant called PortfolioXpert, designed to make Siemens’ complex portfolio more accessible and client conversations more meaningful. Instead of jumping between slides, documents, and spreadsheets, sales representatives could rely on a single interface that guided the flow of a meeting or during conferences.
The assistant combined natural dialogue with structured responses: it starts by gathering context through an onboarding questionnaire, then moves seamlessly into live conversations where it could present overviews, dive into KPIs, highlight operational challenges, and propose opportunities for improvement. Each tab works like a chapter in the conversation: setting the stage, exploring performance, surfacing risks, and ending with actionable recommendations.
Questionnaire
The questionnaire tab acted as onboarding: it asks about industry, company size, revenue, top KPIs, and business priorities. The assistant guided the process step by step, showing progress and setting expectations: answers are then summarized in a clean profile card that the assistant could then use to tailor recommendations.
Overview
The overview tab worked as a dynamic starting point for conversations: instead of static dashboards, the assistant provided a narrative snapshot of operations: production activity, efficiency, energy use, quality, and safety.
Users could then drill into areas of interest: for example, asking for comparisons over time, identifying anomalies, or even switching to voice-based checklists for quick safety confirmations: the design emphasizes clarity and tries giving sales reps a confident way to open discussions.
KPIs
The KPIs tab offers deeper insight into performance metrics: the AI assistant presents key indicators across categories (such as operational efficiency, reliability, sustainability, and cost) in a conversational way. Users could request comparisons (month, quarter, year), explore trends visually with inline charts, or even run simple what-if simulations to understand potential impact. This turned KPI discussions from one-way reporting into an interactive, guided analysis, where the assistant not only delivered numbers but explained their meaning.
Challenges
The Challenges tab helps surface risks and problem areas proactively: instead of only reporting issues, the AI assistant prioritizes them by urgency and potential impact, and offered context such as possible root causes or future projections.
Visual aids like severity tags and trend charts make risks easier to understand at a glance. Also, the AI also encourages dialogue asking the user what to prioritize, offering different mitigation paths, and making the experience collaborative rather than passive.
Opportunities
Finally, the Opportunities tab shifts the conversation to growth and improvement because, here, the assistant suggests potential savings, efficiency gains, or sustainability wins based on the client’s profile.
Opportunities are paired with projected benefits (cost savings, performance improvements, ROI timelines) and visualized with simple charts or summary cards, and the design makes it easy for sales reps to connect client priorities with Siemens’ solutions, creating a clear bridge from conversation to action.
Closing notes
The conversational assistant fundamentally changed how Siemens sales reps engaged with clients: turning long, fragmented discussions into clear, structured, and confident conversations.
Mainly, by reducing the time needed to surface relevant data, this AI tool lets sales representatives focus on building trust rather than managing slides or reports, completely relying on their memory and knowledge.
This shift in the design direction not only improved efficiency, but also boosted customer satisfaction and created new opportunities for lead generation at big events organized by Siemens.
From the team
Jul 23 - Mar 24
Siemens' sales reps needed help gathering customer needs quickly, translating them into structured insights, and recommending tailored portfolio options in real time. We partnered with them to design the PortfolioXpert, an AI conversational tool that began as a digital companion and later evolved into an intelligent chatbot for conferences.
Full breakdown of the case study available on request
Copyright © Mattia Capozzi Sorrentino 2025
Copyright © Mattia Capozzi Sorrentino 2025
Context
Siemens PortfolioXpert was developed to simplify how Siemens sales representatives and customers interact with Siemens’ extensive industrial portfolio. Initially, it served as a structured tool to capture client needs, document conversations, and match solutions. Over time, the tool evolved into a conversational AI assistant, supporting real-time guidance during sales meetings and customer events.
Goal
The main goal was to make Siemens’ complex portfolio easier to navigate, both for sales representatives and for their clients. Conversations with clients were often inconsistent, time-consuming, and focused too much on reporting instead of value. We wanted to change that by creating a tool that could guide meetings in real time, surface the right information instantly, and build more trust between Siemens and their customers.
Impact
Team
I spent time understanding the world of Siemens sales reps and their clients: through 23 interviews, one survey, and two workshops (assumption-mapping workshop and AI ideation workshop), we uncovered a consistent set of challenges:
Conversations with clients were often unstructured.
The vast Siemens portfolio was difficult to navigate in real time.
Sales reps were losing hours on manual reporting instead of building relationships.
Product discovery
User Personas
As part of discovery, I created lightweight personas to keep our focus on the people using this tool every day.
For example, a senior sales representative juggling multiple client meetings in one week needed quick, structured support to stay consistent and credible in front of customers: mapping goals, frustrations, and feature needs helped the team align the team on what really mattered: efficient conversations, simplified portfolio navigation, and clear professional reports.
These personas became practical design anchors, guiding deliverables like journey maps, workshop boards, wireframes, and eventually the conversational prototypes that formed the backbone of the solution.
Ideation
With research insights and personas in place, I moved into ideation: the goal here was to make the assistant feel both conversational and structured, something that could guide a meeting without overwhelming it.
I sketched out early conversation flows, exploring how prompts, quick replies, and structured answers could work together. From there, I built low-fidelity wireframes that simulated real dialogues between sales reps and the AI.
These wireframes let us test different tones, interaction patterns, and layouts quickly: by putting these rough prototypes in front of stakeholders and reps, we could validate what felt natural in a live conversation and refine the flow before investing in high-fidelity design.
PortfolioXpert: a breakdown
The final product was a conversational AI assistant called PortfolioXpert, designed to make Siemens’ complex portfolio more accessible and client conversations more meaningful. Instead of jumping between slides, documents, and spreadsheets, sales representatives could rely on a single interface that guided the flow of a meeting or during conferences.
The assistant combined natural dialogue with structured responses: it starts by gathering context through an onboarding questionnaire, then moves seamlessly into live conversations where it could present overviews, dive into KPIs, highlight operational challenges, and propose opportunities for improvement. Each tab works like a chapter in the conversation: setting the stage, exploring performance, surfacing risks, and ending with actionable recommendations.
Questionnaire
The questionnaire tab acted as onboarding: it asks about industry, company size, revenue, top KPIs, and business priorities. The assistant guided the process step by step, showing progress and setting expectations: answers are then summarized in a clean profile card that the assistant could then use to tailor recommendations.
Overview
The overview tab worked as a dynamic starting point for conversations: instead of static dashboards, the assistant provided a narrative snapshot of operations: production activity, efficiency, energy use, quality, and safety.
Users could then drill into areas of interest: for example, asking for comparisons over time, identifying anomalies, or even switching to voice-based checklists for quick safety confirmations: the design emphasizes clarity and tries giving sales reps a confident way to open discussions.
KPIs
The KPIs tab offers deeper insight into performance metrics: the AI assistant presents key indicators across categories (such as operational efficiency, reliability, sustainability, and cost) in a conversational way. Users could request comparisons (month, quarter, year), explore trends visually with inline charts, or even run simple what-if simulations to understand potential impact. This turned KPI discussions from one-way reporting into an interactive, guided analysis, where the assistant not only delivered numbers but explained their meaning.
Challenges
The Challenges tab helps surface risks and problem areas proactively: instead of only reporting issues, the AI assistant prioritizes them by urgency and potential impact, and offered context such as possible root causes or future projections.
Visual aids like severity tags and trend charts make risks easier to understand at a glance. Also, the AI also encourages dialogue asking the user what to prioritize, offering different mitigation paths, and making the experience collaborative rather than passive.
Opportunities
Finally, the Opportunities tab shifts the conversation to growth and improvement because, here, the assistant suggests potential savings, efficiency gains, or sustainability wins based on the client’s profile.
Opportunities are paired with projected benefits (cost savings, performance improvements, ROI timelines) and visualized with simple charts or summary cards, and the design makes it easy for sales reps to connect client priorities with Siemens’ solutions, creating a clear bridge from conversation to action.
Closing notes
The conversational assistant fundamentally changed how Siemens sales reps engaged with clients: turning long, fragmented discussions into clear, structured, and confident conversations.
Mainly, by reducing the time needed to surface relevant data, this AI tool lets sales representatives focus on building trust rather than managing slides or reports, completely relying on their memory and knowledge.
This shift in the design direction not only improved efficiency, but also boosted customer satisfaction and created new opportunities for lead generation at big events organized by Siemens.
From the team