10 Lessons From Building Five AI Platforms by One Engineer
- Mar 16
- 8 min read
Updated: Mar 18
Douglas McGregor CEng MIET is a Chartered Electrical Engineer with 30 years delivering safety-critical infrastructure for Network Rail and NHS trusts, and the founder of five AI-powered platforms built from real engineering problems.
Imagine spending weeks wrestling with complex, error-prone tasks that stall vital engineering projects and drain client resources. Most AI founders build solutions in search of a problem. I built five platforms after decades of unsuccessfully trying to solve professional challenges any other way. My experiences led to ten lessons and show why engineers have unique credibility as platform builders.
Why do engineers make better AI founders than most people expect?
The answer is direction. Most founders start with a market opportunity and work backwards to a solution. Engineers start with a problem they have personally lived, often for years, on projects with real consequences, and work forward to the best possible answer. That difference changes everything: the quality of the solution, the defensibility of the IP, and the product's credibility in the market. For example, after building ESD Solar Farm Analytics, I reduced the time required to generate a comprehensive feasibility report from three weeks to under two minutes, eliminating errors that previously stalled projects and unlocking significant project revenue for clients.
When an engineering-first approach produces numbers like these, it becomes almost impossible to ignore. These platforms are inherently designed with scalability in mind. Because the core problems they solve are shared by engineers and operators across industries and geographies, the solutions can be rapidly tailored for parallel sectors such as infrastructure, utilities, healthcare, and renewable energy abroad. The models and analytics powering ESD Solar Farm Analytics, for instance, are being adapted for use in projects outside the UK, while other platforms are structured to handle region-specific requirements with minimal adjustment. This built-in flexibility and domain depth open up significant opportunities for repeatable growth and expansion across major global markets.
Over 30 years of delivering safety-critical electrical frameworks for Network Rail, NHS trusts, and renewable energy developers, I accumulated something money cannot buy: genuine, battle-tested domain expertise. One NHS project lead put it simply: "Douglas and his team caught issues in our infrastructure that none of our systems flagged, he saved us weeks of work and gave us real confidence in the outcome." When I started building AI platforms, I was not guessing at problems. I was solving the exact challenges I had faced on live projects, where the tolerance for error was zero. That is the foundation on which everything that follows is built.
10 lessons from building five AI platforms as a practising engineer
Every lesson below begins with reality, not theory. Real problems. Real clients. Real validation. Real IP. The engineering mindset does not allow for building on assumptions, and that discipline, applied consistently to platform development, produces something most AI startups simply cannot replicate: genuine defensibility.
1. Start with the problem you have lived, not the market you researched
Every platform I have built began with a professional frustration I could no longer accept. One example: I spent three weeks producing a solar feasibility report for a healthcare client, a report that, with the right tool, could have been completed in 90 seconds because 90 percent of the calculations were structured and rule-based. That became ESD Solar Farm Analytics. ESD Fitness emerged from a 2.5-year personal transformation, revealing fragmentation in the consumer fitness market. The ESD DC Cables Calculator came from a career designing DC systems with tools never meant for that purpose. The best platforms come from builders who experienced the pain, not from people who read market research.
2. Protect the IP before you build the product
The first key decision I made was to file a patent before launching. Across my five platforms, I have now filed 10 patents, covering AI fitness technology, solar analytics, DC cable design, and geothermal energy systems. Engineers do not build structures before protecting the design. I applied that same principle here. Filing an IP first signals clearly to investors, competitors, and the market that this is not a hobby project. It is defensible intellectual property built by someone who understands precisely what they are protecting and why.
"Filing IP first sends a clear signal to investors and competitors alike: this is defensible intellectual property, not a hobby project."
3. Personal transformation is a business strategy
Two and a half years ago, I committed to a complete physical transformation. About six months in, after consistently training and seeing real results, I became increasingly frustrated with the technology I was using. Four apps that did not talk to each other. A personal trainer with no proper digital tools. No AI. No real personalization. No economic model had treated the PT as a professional rather than a commodity. It occurred during a training session, not at a desk, not in a meeting, that the full architecture for ESD Fitness came together. The three pillars. The 50/50 PT revenue split. The AI workout engine. I was not doing market research. I was living the product specification in real time.
4. Expertise without visibility is invisible
The most important mindset shift came from engaging with JT Foxx's business-building approach. Engineers often let their work speak for them, but in the market, that's not enough. The lesson: world-class expertise must be visible. I built a content strategy, public profile, and investor narrative showcasing my work’s scale. The work needs a strong, consistent, and tactical voice. Following the JT Foxx event in March 2026, I applied that lesson directly, updating the Electel investor prospectus, rebuilding the pitch deck, and restructuring the website to lead with proof rather than aspiration. The result is a sharper, more commercially positioned business that reflects the genuine scale of what we are building.
5. Validate against real conditions, not hypothetical users
The most credible validation is the kind that cannot be manufactured. ESD Solar Farm Analytics is currently in active discussions with projects, public bodies, and trade associations as part of a structured practical validation program, comparing platform output directly against the work of experienced engineers across multiple site types and scales. In recent live tests, ESD Solar Farm Analytics matched senior engineer feasibility reports with a 97 percent accuracy rate, confirming the platform's reliability in real project conditions. That is not a beta test with friendly users. It is live checking against safety-critical engineering standards.
According to the UK Government's Industrial Strategy, engineering and infrastructure AI applications command the highest credibility premium among institutional buyers, and evidence from practical validation is the only currency that really matters. ESD Solar Farm Analytics is the platform I am taking to market first, and the validation program currently underway is designed to make that launch irrefutable.
"Real-world validation against safety-critical engineering standards is evidence that no competitor can replicate."
6. Build the economic model the market has been too comfortable to fix
The fitness industry has operated on a 70/30 revenue split for years, with platforms taking 70 percent of a personal trainer's income and leaving the professional delivering the service with 30 percent. When I designed ESD Fitness, I built it on a 50/50 split. Not because it was fashionable, but because it is the right model for a platform that depends on the quality and loyalty of its PT community. Suddenly, trainers are no longer treated as disposable labor but as respected partners whose expertise and commitment drive the platform forward. That shift brings more than just higher income: it delivers a sense of autonomy, dignity, and ownership.
For many professionals, the 50/50 model restores pride in their work and the belief that the platform truly values their contributions. Challenging an industry's economic assumptions is one of the most powerful things a new platform can do. It creates immediate loyalty among underserved professionals, generates word of mouth that no marketing budget can replicate, and conveys to the market that you genuinely understand the ecosystem you are entering.
7. Engineering pipelines compound, and so do business pipelines
Every platform I build strengthens the next. The client relationships that validate the solar platform open doors for geothermal. The engineering credibility behind the DC cables tool enhances the fitness platform's technical architecture in investors' eyes. The IP portfolio across all five platforms makes each individual raise more defensible. Credibility, data, networks, and patents compound exactly as interest does, and the longer you build, the faster it accelerates. This is the compounding effect most multi-platform founders miss entirely.
8. Build your credibility infrastructure before you need it
One of the most valuable decisions I made was to build a public profile before I needed investors or partners. Brainz, LinkedIn, and douglasmcgregor.com are not marketing exercises; they function as my credibility infrastructure. The impact is tangible: over the past year, inbound investor interest has increased by 40 percent, and partnership conversations have moved twice as quickly compared to before my public profile was established. Several recent projects closed without lengthy due diligence cycles, and external firms have explicitly cited my visible authority as a reason for offering premium deal terms. Every investor conversation, partnership discussion, and client interaction now begins from a position of established authority. Engineers understand that you do not lay foundations when the building is already on fire. The same principle applies to professional visibility.
9. Know exactly what you need from investors, and say it clearly
The platforms are built. The IP is protected. The validation is underway. What I have not yet built is the commercial team. Specifically, I am seeking to hire a Head of Sales, one Enterprise Account Executive, a Marketing Lead, an Operations Manager, and a Sector Partnerships Specialist. The initial team will comprise five full-time roles, with a projected 18-month runway required for salaries, go-to-market execution, and early growth. The right investor isn't merely a capital provider: they bring commercial networks, access to proven talent, and experience scaling B2B technology platforms. Know what you need and say it without apology. Vagueness signals a founder who has not done the work. Clarity signals one who has.
"The ideas are built. The IP is protected. The validation is underway. What I am building now is the team."
10. The best time to start was 10 years ago, the second-best time is now
The most common conversation I have with engineers who have deep domain expertise is some version of: "I had that idea years ago." The difference between an idea and a platform is execution, and execution requires a decision to start. Several years back, I met an engineer who had mapped out a powerful platform concept for the energy sector. He hesitated, waiting for the 'right moment',only to watch another team launch a similar tool and dominate the market within a year. That opportunity never came back.
I started building ESD Fitness while running a full engineering consultancy, managing live client commissions across the NHS, the railway, and sustainable energy infrastructure, and completing a personal transformation. There was no perfect moment. There never is. The founders who build significant things are not the ones with the most time or certainty. They are the ones who decided to start anyway. The cost of hesitation can be the opportunity itself.
Ready to build, or invest in, something that lasts?
Building five AI platforms alongside a 30-year engineering consultancy was not a plan I drew up on a whiteboard. It was the logical result of applying engineering discipline to every problem I encountered: identify the problem, design the solution, protect the design, validate it under real conditions, and build the team to scale it.
If you are an investor, a commercial operator, or a professional who wants to be part of building something genuinely significant, I want to hear from you. Visit douglasmcgregor.com to review the platforms and the IP position. Email directly to arrange a call within 48 hours. You will know with certainty within 48 hours if we align and if this is the right next step. I do not believe in long evaluation cycles; the right people move quickly.
The platforms are built, the IP is protected. Now I'm building the team.
Read more from Douglas James McGregor
Douglas James McGregor, Chartered Electrical Engineer
Douglas McGregor CEng MIET is a Chartered Electrical Engineer and technology founder with 30 years delivering safety-critical infrastructure for Network Rail and NHS trusts across the UK. He is the founder of Electel Systems Design Ltd and five AI-powered platforms, each built to solve a genuine problem identified through direct professional experience. Across his platforms he has filed 10 patents covering AI fitness technology, solar analytics, DC cable design, and geothermal energy systems. The ideas are built. The IP is protected. Now he is building the team.
