Charting a new course. Storm’s vision for the AI era

At Storm we believe AI is not just an incremental advance in tech like cloud, mobile or social but instead represents an opportunity to completely rewire how organisations operate and create value.

Personally, it’s the most exciting time to be involved in technology since the early 2000s and Storm are leaning into this new shift. Here is a breakdown of what we are doing:

Storm as an AI first organisation

At Storm embracing AI effectively is now a fundamental expectation of everyone in the company. AI is rapidly becoming an indispensable tool across all disciplines of the business and we expect its significance will only escalate.

To empower our teams every Storm employee now has a budget for two paid AI tools. Within a few months we have a 90% take up across the company using tools such as ChatGPT Team, Claude Team, Gemini, Cursor, GitHub co-pilot, Eleven Labs, JetBrains Rider and MS co-pilot. We have also completed risk and compliance reviews across all these tools to align with our ISO 42001 standard.

We are running regular internal show and tell sessions on use of AI internally, actively encouraging sharing of best practice on internal channels and generating monthly evaluation reports on time saved per tool and per team which is yielding some incredible data.

We’ve now established an internal team focused on developing action-based AI within the company. This team is identifying high value, technically feasible use cases to drive automation across key internal workflows in sales, marketing, HR and customer support utilising platforms like Co-Pilot Studio.

Talent acquisition strategy

Storm boasts a world-class, multi-disciplinary team of seasoned digital professionals spanning software development, cloud platform and data engineering, data science and analytics, agile product development and delivery, testing, UX/UI design, and digital strategy.

Whilst we're committed to augmenting our experienced team with cutting-edge AI tools, we are also passionately investing in the next generation of AI-native talent. These new AI tools significantly reduce the time and effort required for senior professionals to train and mentor emerging talent. The equation on bringing on junior talent has fundamentally changed.

As a result, we are actively recruiting recent graduates and early-career professionals across two key role types:

• AI Engineer
• AI Consultant

AI Consultants will leverage an internal framework to identify optimal AI use cases, map current workflows, assess existing technology stacks and evaluate change management needs. Drawing from our experience in clinical trials, we've developed robust evaluation frameworks to assess AI system performance, monitor model accuracy, detect data drift, and meticulously measure and report on business impact.

Selecting the right use case

More details on the roles and how to enquire here:
https://stormid.com/careers/

Agentic workflows

Knowledge based AI where individuals use tools to improve personal productivity is the most advanced in terms of adoption today, but we soon start to see move towards action-based AI or what is being termed AI agents.

It’s likely AI agents will be as popular as mobile apps and as the “Agentic Web” evolves it will start to displace many of these mobile apps and certain SaaS products.

These agents are designed to make decisions and perform tasks on behalf of individuals and organisations, interfacing with internal or external systems with minimal human oversight.

At Storm we have spent the last year building an Agentic Workflow Engine which is designed to accelerate AI adoption and help our customers transition from small scale proofs of concept to deploying AI solutions at scale.

The workflow engine supports rapid creation of agentic workflows by customising existing templates or building new agentic workflows from a suite of core components. We deploy these workflows directly to the customer’s own cloud tenancy, integrate them with internal data and systems and provide human in the loop audit trails for secure, low-risk adoption.

In the short term we can use AI Agents to add a layer of intelligence on top of existing systems, in the medium to longer term there is an opportunity to reimagine a service to unlock AI’s full potential.

The types of agents we are working on include a combination of the following:

• Assessing applications, claims or data against clear guidelines
• Classifying service requests according to a set of rules
• Validating data submissions for completeness or compliance
• Prioritising tasks or cases based on criteria
• Summarising large volumes of text
• Generating routine reports or communications
• Transcribing audio or voice interactions

Building a workflow agent

Find out more about Agentic Workflow Engine:
https://stormid.com/agentic-workflow-engine/

Ready to explore how your organisation can harness AI? Consider one of our AI Sprints:
https://stormid.com/data-and-ai-sprints/

For Public Sector organisations don’t miss the AI Challenge:
https://stormid.com/futurescot-ai-challenge/

Sovereign AI

While we collaborate closely with public cloud providers like Microsoft Azure, we recognise that certain AI workloads demand deployment within your own network on local systems. For scenarios where data sensitivity, privacy concerns or specific regulations necessitates that data remains off public cloud providers, Storm is partnering with local data centres to offer Sovereign or Private AI options.

With Sovereign AI, we securely deploy AI agents or AI systems within your own network, connected directly to your data using open-source models. This is underpinned by a robust compliance framework covering data residency, data governance, regular compliance checks and a comprehensive range of security controls.

Voice

The way we interact with technology is going to change over a period measured in 2-3 years rather than five or more. The word error rate for European language speech to text transcription has over the last few months dropped below 7% - a word error rate of under 5% is seen as human level performance. Latency for voice interactions with machines has also reduced significantly over the last 12 months with average response times now around 300 milliseconds – a similar response time to most human conversations.

The potential of using voice as a data input is huge. Field or mobile workers can save vast amount of time dictating directly into case management systems. Voice could also be the major unlock for healthcare data interoperability, freeing clinicians from repetitive and tedious data entry into poorly designed EHRs and other clinical systems? It can also offers a solution to engaging digitally with those with accessibility needs or lower digital literacy.

Conclusion

AI progress is accelerating with advances not just in the models themselves but also in the underlying chips and data centres. It’s highly probable that within the next two years we will all have access to a doctor, teacher, lawyer in our pockets – largely free of charge.

AI will drive a lot of today’s mundane admin tasks to zero and create space for entirely new professions and new AI native products and services. The opportunity to apply this across a range of sectors including our over-stretched public services is huge.

So, if you have a problem to solve, an idea to explore or a service you want to transform with AI, please get in touch.

It’s time to build.