Resources to help you build, manage,
and evolve your architecture
Dive into playbooks, reports, and insights that help teams cut through vendor bias and build a smarter tech stack that's ready for what's next.
Filter by
Showing
0
of
0
items
Blog
May 11, 2026
0
min read
The Technology Pillar
Most companies don't have an AI stack. They have AI features bolted onto yesterday's processes.
Sean King
CoFounder of Dragonfly
May 11, 2026
0
min read
Sean King
CoFounder of Dragonfly
The Technology Pillar
Everyone is talking about AI. Copilots, agents, autonomous workflows. But when we sit down with the founders, CTOs and ops leaders actually running fast-scaling companies, the reality looks nothing like the LinkedIn posts.
People in over 90 percent of companies are using AI tools at work today. The trouble is that almost none of those companies can answer the only question that actually matters: where exactly is my business today, and what's the next move?
That's the job of the Dragonfly AI Readiness Framework. It evaluates a business not by how many tools it owns, but by how much human effort is still required to move work from one step to the next inside its core processes. Across that mapping work, in sales, support, engineering, HR, finance and marketing, one pattern repeats: most scaling companies are stuck somewhere between digitised and automated, nowhere near ready to unlock what AI can actually do for them.
Three pillars, one honest diagnostic
The framework looks at readiness through three pillars.
Technology asks whether the tools at each stage of your core processes are actually doing the work, or whether your team is silently filling the gaps between them with copy-paste, side spreadsheets and end-of-day catch-ups.
People is about what happens after the tools are in. If you've shipped AI to your team and watched most of them drift back to their old workflows within weeks, the People pillar explains why that isn't an adoption problem at all, and what has to shift underneath for any of the tools to actually stick.
Strategy is what holds the other two together. If your AI strategy was greeted with applause at the offsite but you can't name a single thing it has changed in the business since, the Strategy pillar lays out the decisions any real strategy has to commit to, the ones most leadership teams quietly skip.
Inside the Technology paper, a first look
This summary covers paper one of three, the Technology pillar. It comes first for a reason. People won't adopt tools that don't fit how their work actually flows, and no strategy can compensate for a stack that's quietly held together by human effort. Before you can change the work, you have to see it clearly. That's what the Technology pillar is for.
Within Technology, every core process in your business sits at one of five readiness levels. These aren't marketing tiers. They represent a fundamental evolution in what you're asking your software to do, and they determine the ceiling of what AI can deliver for you.
- Level 1, Analog. Whiteboards, spreadsheets, and people remembering things in their heads.
- Level 2, Digitised. Systems of record. CRMs, ERPs and helpdesks that store information but don't move it.
- Level 3, Operational. Rules and macros stitching tools together. Efficient, but fragile.
- Level 4, Augmented. Software that thinks with your team. Predictive insights, copilots, context-aware workflows.
- Level 5, Autonomous. Software that runs end to end. You set the outcome, the system does the work, and only loops in humans when judgement is actually required.
AI adds value at every level. But the magnitude compounds as you climb. A company stuck at Level 2 is leaving most of AI's potential on the table, not because they lack ambition, but because their infrastructure can't support it.
The trap: everyone thinks they're at Level 4
Almost every leadership team we sit with believes they're already operating at an advanced level. They have a CRM. They have dashboards. They've rolled out a copilot, maybe experimented with agents. So they assume they're AI-ready.
But when we map the real flow of work, the picture changes fast. Data is still being copied between systems by hand. Decisions still live in people's heads. Automation exists, but it's brittle. A patchwork of if/then rules duct-taped across tools that were never designed to work together.
That's Level 2 and Level 3 behaviour wearing a Level 4 badge. It's also exactly where most AI projects quietly fail.
AI doesn't fix broken processes. It exposes them.
Layering intelligence on top of fragmentation doesn't transform anything. It just makes the cracks faster and more visible. The real work isn't choosing better AI. It's seeing your processes clearly enough to know which ones can actually carry intelligence, which ones need to be redesigned first, and which ones are being held together by nothing but human patience.
That diagnostic, applied to your business, is what the rest of the Technology paper delivers.
What's in the full paper
What you've just read is the framework in outline. It's the part that fits in a website summary. The paper itself is where the framework becomes a tool you can actually use, and that's the part that changes how you spend the next twelve months. The summary makes the case. The paper makes the difference. Inside, you'll find:
- The full anatomy of each of the five levels: what AI can and can't realistically do at each, why the ROI equation flips between Level 3 and Level 4, and where most companies actually sit, which is rarely where they think they sit.
- The decision framework for when Level 5 is the right call and when chasing it is a mistake. Moving every process to autonomy isn't the goal, and treating it as one is as misguided as buying tools without a strategy.
- Worked examples for outbound sales and customer support, mapping the exact stages, tools and metrics that shift as you climb the framework.
- How Dragonfly's AI Readiness Platform turns the framework into a live benchmark of your specific stack against the best alternatives in the market, so you stop guessing and start comparing.
If you stop reading at the summary, you'll have a sharper way of talking about AI readiness. If you read the paper, you'll have a sharper way of running the business.
The next generation of companies won't win by owning more software. They'll win by evolving faster than their market, and that journey starts with brutal clarity about how your business actually runs today.
Download the full paper to see where yours stands.
.avif)
News
February 14, 2026
0
min read
Dragonfly raises £2.6 million pre-seed to help businesses get ahead with smarter decisions about software
Dragonfly, the world’s most powerful software discovery platform, today announced the close of a £2.6M pre-seed round alongside the public launch of its conversational AI tool. Founded in December 2024 by Zego alumni Sean King and Sven Sabas, the round was led by Episode 1, joined by Dreamcraft and Portfolio Ventures, with angels including QuantumBlack founder and CTO Sam Bourton, and Bolt founder and CEO Markus Villig.
Sean King & Sven Sabas
Dragonfly Founders
February 14, 2026
0
min read
Sean King & Sven Sabas
Dragonfly Founders
Dragonfly raises £2.6 million pre-seed to help businesses get ahead with smarter decisions about software
The number of software tools has exploded in recent years. Tech users no longer find themselves facing a decision between an incumbent provider and a challenger; today, there may be hundreds of options to consider for any given job. The rise of AI and vibe-coding has rapidly accelerated the development of new applications, but substantially increased the time needed to evaluate the security, reliability and interoperability of different technologies. Expert advice on business architecture - from systems architects or management consultants - is inaccessible to the majority due to high costs, and the pace of innovation makes it impossible for anyone to stay up-to-date.
Dragonfly aims to make the expertise of a solutions architect instantly accessible to everyone, from individual professionals to enterprise leaders. The startup has compiled the world’s largest catalogue of software tools and their capabilities, currently standing at over 250,000 products, from a range of trusted data sources. This dataset feeds into a variety of AI-powered features, which will help users find the right tech stacks in seconds, rather than months.
Today, Dragonfly has made its proprietary conversational AI tool available to the public. Instead of scrolling through marketplaces filled with outdated and incomplete information, or sitting through lengthy sales demos, users will be able to ask Dragonfly questions; within seconds, Dragonfly will recommend a list of suitable products, ranked by relevance, with supporting explanations.
The tool is designed to cater for simple queries in plain language, such as "what’s the best project management software for a team of five that integrates with Google Calendar and has a free plan," and more complex prompts, for example "we need to build a secure, compliant technology architecture to support a new financial services product for our European customers; we need recommendations for tools that can handle real-time data ingestion, are GDPR-compliant, and integrate with our existing AWS infrastructure, and we'd also like to see how these tools would work together in a single blueprint.
From late 2025, Dragonfly is set to roll out its enterprise offering to help organisations build, manage and evolve their tech stacks. Businesses will be able to map out their software with the aid of Dragonfly’s ‘digital fingerprinting’ technology, which captures a blueprint of their business system architecture. With this detailed context, Dragonfly will supply insights and software recommendations tailored to each company’s environment, goals and growth stage.
King and Sabas have assembled a team of 12 with several former colleagues from Zego, Peppy and Permutive joining the founders in their new venture. The pre-seed funds will be used to develop a range of new features in support of the startup’s mission to build a comprehensive ‘Automated Solutions Architect’, further enrich Dragonfly’s data, and recruit select new hires as needed.
Hector Mason, General Partner at Episode 1, said: "Before Dragonfly, it was almost impossible to gather enough context about software tools to truly understand their capabilities, and whether they fit seamlessly within an existing stack. By solving this critical problem, made even more urgent with the rapid adoption of AI, Dragonfly has the potential to make every company on the planet operate more effectively.
Sam Bourton, co-founder and CTO at QuantumBlack, said: “I first met Sean and Sven after having spent three days with 150 CIOs and CTOs, witnessing firsthand the mounting complexity of the modern tech and data landscape. The explosion of GenAI and agentic systems has amplified the problem - leaders face a firehose of requests for new AI tools, with a thousand flowers blooming across their organisations, each demanding attention and resources. Taking stock of their existing tech stack, evaluating options, and managing security, dependencies, and integrations has never been harder.
To try Dragonfly for free, visit www.askdragonfly.com or see our social media channels for latest updates.
Subscribe to our newsletter
Catch the latest trends and insights straight in your inbox.
We’re happy to have you with us