Rethinking innovation: How AI is reshaping value creation

Tagging thousands of products quickly and efficiently is a real challenge. We helped a retail client tackle this challenge by building an AI system that analyses product images and applies accurate tags. Previously this process was manual, delaying products reaching the e-commerce platform. Today, the AI recognises popular characters, learns from previous tags and can identify new ones. With human oversight ensuring accuracy, the system now handles 95% of tagging, saving time and accelerating launches online.

An international human services provider needed a more efficient and consistent way to document the hundreds of face-to-face meetings its case workers conduct every day. Producing detailed written summaries manually was time-consuming, created inconsistencies and reduced the time available for direct community engagement.

We helped the company build a purpose-built AI summarisation tool that captures meeting audio, creates a transcript and generates a structured summary tailored to the type of meeting. Case workers can then review and refine the summary before it is finalised, ensuring professional judgement remains part of the process.

By using AI to support meeting documentation, the organisation can reduce manual administration, improve consistency and give case workers more time to focus on supporting the community. The solution also provides greater control over data, governance and compliance compared with standard transcription tools.

AI platforms are changing how consumers find products, compare brands and make buying decisions. While referral traffic from tools such as ChatGPT, Gemini and Perplexity is still developing, it only shows part of the picture.

The bigger challenge for ecommerce businesses is understanding what happens before a customer reaches their website. Which prompts are shoppers using? Which brands are being recommended? How often does your brand appear? And how does that compare with your competitors?

That’s why we’ve developed an LLM Visibility Report designed to help businesses understand their presence across AI-driven discovery journeys. It identifies relevant prompts, tracks brand recommendations, measures visibility and sentiment, and compares performance against competitors.

From there, our GEO audit provides clear recommendations on how to improve visibility, from strengthening website content and schema to improving how AI platforms access, understand and reference your brand. This gives ecommerce businesses a clearer view of where they are performing well, where they are less visible and what actions they can take to improve visibility as AI becomes a growing part of product and brand discovery.

How do you turn AI ambition into practical use cases that deliver value across multiple business units?

For one multi-entity group providing financial, technology, creative and fleet management services, there was strong ambition to apply AI across finance, operations and sales. But ideas were fragmented, processes remained manual and systems were highly siloed. Through targeted user studies and envisioning workshops, we worked with the group to identify where AI could deliver near-term value and designed three agentic AI use cases focused on core business processes.

These included an agent to support invoice processing and vendor onboarding, an operations agent to improve ticket handling, technician dispatch and billing, and a sales agent to support lead qualification, technical inputs and quote generation. Together, these use cases helped reduce manual effort, improve process consistency and remove bottlenecks across key areas of the business. It also created a reusable foundation for extending agentic AI into additional processes and business units.

How do you make valuable knowledge easier to find when it’s buried in several historical documents?

For one online marketplace for residential rentals, a large archive of historical house and plan drawings contained important information about room layouts, areas, orientation and architectural details. However, this knowledge wasn’t structured, making it difficult to reuse in sales and early planning conversations.

We helped develop an AI-driven property assistant that analyses existing drawing material and turns it into structured data. The solution can identify information such as room types, sizes, layout characteristics, access details and facade features, making it easier for teams to search, compare and explore existing properties and projects.

By creating a structured knowledge layer on top of the existing archive, the solution reduces preparation time, improves the quality of customer dialogue and makes specialist knowledge easier to access. It also reduces reliance on individual experience, helping teams reuse valuable information earlier in the sales and planning process.

How much time is your team spending registering hours, coordinating invoice drafts and checking for errors before invoices can be sent?

For one construction and property development company, the invoicing process involved several manual steps and handovers. This created unnecessary admin, added pressure on resources and increased the risk of errors in invoice drafts.

The solution? An AI agent that registers hours, connects them to the right project information and generates invoice drafts automatically. These drafts are then validated against contracts and documentation before being reviewed.

The result is a simpler, more accurate invoicing process. The chat-based interface makes time registration easier for employees without the need for training, while automated validation reduces errors and manual quality checks. By implementing the AI agent solution, the company can significantly reduce invoicing time while generating cost savings every year.