Case Studies

The Problem

Creative and communications professionals are often skeptical of AI, and for good reason. Generic AI outputs can undermine brand voice, introduce inaccuracies, and produce work that feels hollow. MTP needed AI embedded into their actual production process: from crisis response to strategic planning to high-volume content creation. The goal was repeatable, workflow-level integration, not one-off experimentation.

Our Role

Designer and facilitator of the series. Topics were chosen each month together with MTP's CEO.

What We Did

Second Mind Solutions gave MTP leadership a menu of AI topics mapped to the agency's core production functions, from crisis response to strategic planning to high-volume content. Each month, Second Mind and MTP's CEO selected the next topic together, prioritizing the functions with the most immediate relevance. Each session was then built around realistic scenarios from that function, with live demonstrations and hands-on application, so the team could see how AI applied to the kinds of tasks they handle every day.

What Was Delivered

• Session 1: Crisis communications workflow, rapid response using AI prompt chaining and persona layering
• Session 2: Strategic planning workflow, using Gemini for scenario planning, message testing, and competitive analysis
• Session 3: Visual content production workflow: image generation, mockups, and short-form video
• Session 4: High-volume creative production, integrated workflows across the full Google Workspace AI ecosystem
• A proposal for a custom AI Studio Day to convert learned patterns into permanent, MTP-specific production workflows

What Changed or Was Learned

Across the series, MTP's team worked through applied AI patterns spanning crisis communications, strategic planning, visual production, and high-volume creative output, with prompt frameworks and techniques they could carry into their own work. The series gave leadership a clearer view of where AI fit across the agency's production functions, and led to a proposal (a custom AI Studio Day) for turning these patterns into permanent, MTP-specific workflows.

The Problem

BNET had deployed enterprise AI licenses to approximately 150 staff across customer service, operations, marketing, and HR, but leadership could not see what it was producing. Some employees experimented on their own; most didn't know where to start. Without a structured framework connecting AI capability to actual job functions, the tools sat underused.

Our Role

Designer and facilitator of the training program, including the frameworks, the live session, and follow-up support.

What We Did

Second Mind Solutions mapped AI capability to role-specific workflows across three high-value operational functions (task automation, strategic decision-making, and quality assurance) and built structured, job-relevant frameworks employees could apply immediately. A live session delivered across time zones used interactive polling and role-based demonstrations to make the frameworks concrete and actionable. A post-session innovation challenge and optional office hours gave staff a way to keep applying the frameworks after the session.

What Was Delivered

• Role-specific AI workflow frameworks across customer service, operations, and communications functions
• A practical AI toolkit organized around BNET's three core operational use cases
• A live implementation session for ~150 employees delivered across time zones
• A post-engagement innovation challenge to keep staff applying the frameworks
• Group office hours to address team-specific workflow questions
• A roadmap of advanced follow-on engagements for continued operational integration

What Changed or Was Learned

Staff received role-specific workflow tools and a framework for applying AI to their actual job functions, not abstract AI concepts. Leadership received a structured roadmap for deepening adoption and for measuring progress across the organization.

The Problem

Solopreneurs running every function of their business face a distinct operational problem: they can't afford to work inefficiently, but they also can't afford to adopt tools that don't deliver immediate, reliable value. Generic AI programs rarely address the specific decisions and workflows this group navigates daily, and they often risk homogenizing the voice and identity that makes their businesses distinctive. The program needed to be function-specific, immediately applicable, and protective of each owner's authentic brand.

Our Role

Designer and trainer for all three tracks, delivered through a partnership with a National Urban League initiative.

What We Did

Second Mind Solutions built three workflow-specific AI deployment frameworks, each targeting one of the highest-friction operational areas for independent business owners: financial operations and contract analysis, content production, and project and client management. Rather than covering AI broadly, each track was built around the actual decisions and constraints these entrepreneurs face, with AI deployed as a precision tool for specific, high-stakes functions.

What Was Delivered

• Track 1. Financial & Strategic Operations: AI workflows for contract analysis, financial scenario modeling, complex document summarization, and building custom AI assistants tailored to each business
• Track 2. Content & Communications: A structured production workflow for multi-channel content with built-in brand voice guardrails, designed to keep output on-brand with less manual rewriting
• Track 3. Project & Client Management: An AI-powered project management framework for scoping work, mapping stakeholders, identifying risks, and building repeatable delivery systems

What Changed or Was Learned

Participants left with AI workflows mapped to three high-friction areas of running a business solo, designed to reduce manual work without requiring technical expertise or outside support.

The Problem

An earlier survey had surfaced friction in existing banking apps and heavy resistance to using voice for financial tasks, but it could not say why, or what to do about it. The team needed decision-grade evidence on usability, trust, and adoption to guide a go or no-go call, MVP scope, and go-to-market.

Our Role

Qualitative research lead: discussion guide, interviews, analysis, and the readout.

What We Did

Second Mind redesigned the discussion guide, then ran one-hour interviews in which participants worked through six voice-command prototype scenarios while live survey results informed the probing. Transcripts were coded with a custom-built qualitative analysis GPT, with manual spot-coding for quality control, and cross-referenced against the survey data throughout.

What Was Delivered

• A findings deck mapping each insight to product, trust, monetization, and adoption decisions
• A Kano-style feature framework separating must-haves, satisfiers, and delighters
• A quantified view of adoption drivers and barriers, combining survey and interview evidence
• Willingness-to-pay analysis and a recommended entry model
• A live readout for the development, design, and leadership teams

What Was Learned

The research recommended reframing the product from "voice-first" to a multi-modal wallet with text and adaptive UI at the center and voice as an optional layer, so the build would not rest on its riskiest assumption. It recommended MVP priorities around reliability, confirmations, and user control, and a freemium entry model until trust is established.

The Problem

Insurers traditionally target customers by demographics. The model bet instead that behavior and risk preference predict buying propensity better than personal details, which also reduces reliance on personal data. That bet had to be validated, calibrated for Australian consumers, and turned into something a sales and marketing team could use.

Our Role

Research lead for the qualitative phase and designer of the quantitative validation instrument.

What We Did

Participants self-identified against the model's risk personas, and their perceptions, opinions, behaviors, and attitudes were mapped against that choice. Projective exercises and testing of the tool's digital assets surfaced the behaviors behind each profile and the language that moved each one.

What Was Delivered

• Risk personas recalibrated for the Australian market
• A risk-profile quiz design framework built on observed behavioral anchors
• A profile-to-product propensity mapping
• A conversion and communications strategy for each profile
• A quantitative survey instrument to validate the findings at scale

What Was Learned

In the interviews, risk preference was a stronger signal of insurance propensity than life stage, and some of the most prudent, research-driven prospects were younger than traditional targeting assumed. The studio came away with a recalibrated scoring framework and a treatment plan it could build into the product.