Ash Oldershaw
Software Manager at Naked Energy
Building practical software for complex problems in climate tech and IoT
Building software in climate tech
When the company's main product isn't software and you have a business focused on sales, product performance, and scaling its operations, you have to really understand the whole business to make an impact.
Read more →
Most of my previous jobs have been in companies where software was not the end goal, but it was the enabler that made the company a huge differentiator. As such you need to understand what it is the company offers well enough to be able to sell it yourself. The unique challenges of climate tech are not simply competition between different renewable technologies. Naked Energy's technology complements most heating technologies really well, but how do we build a software offering that communicates that value, and understands the relationship between Virtu collectors, heat pumps, biomass boilers, CHP and good old natural gas? Solar thermal is a niche technology right now, and a lot of what we have to do is actually educate people in what our technology is, and what its pros and cons are. PV is relatively simple to simulate, design and install, whereas solar thermal requires fluid simulations, piping design, plant room simulation, and installation mechanics that make it much more difficult to design by hand. We have made building a detailed desktop design and feasibility study even easier than it is for PV-based systems, with a high quality, pitch ready output.
Making impact under constraints
When you have autonomy but limited resources, building the right thing at the wrong time has a real cost.
Read more →
Do we help the sales team close deals faster? Do we help the operations and maintenance team work more smoothly and provide better insights into the performance of collectors? Do we help the engineering team make designs more quickly? When every part of the company could use some software support but your team's headcount is three and you have autonomy as a team, understanding the pain points of each team and the company's big-picture priorities mean we know when we're on the right track. When the business needs have changed, we have changed our approach to make sure that our output is being used in the areas it's needed the most, and the impact we have is the highest. Part of the investment we have made into this is by using AI as a force-multiplier for the developers we have in the team. Building an AI-enhanced software development lifecycle has given our team the best ROI of any change to our methodology, library choice or tool selection.
Building a product trusted across markets
Single-locale software development is fine when you're selling to one country or solving for one team. Making impact across the world means you need to manage content and context separately, and bring together a coalition of stakeholders.
Successfully integrating external development teams
Managing software development for a project that was being developed externally knowing it would come in-house has unique challenges. Thankfully we managed to avoid them with day-one integration.
Scaling development with AI agents
In 2023 we wrote every line of code ourselves. In 2026 we write almost none, and we do it safely. How do we move faster than ever without breaking things we can't afford to? I've made some bad mistakes and learned some hard lessons, re-learning how to make software is tough.
Why portfolio sites are changing
When the cost of execution is so low, portfolio projects become less about what you are capable of doing and more about what you are capable of solving. AI summaries mean you need to focus on the content of your site more than the styling of it to get noticed.
This site is an agentic artifact
This website was built with agentic AI because the fastest moving teams are using AI to do the things AI is good at, and focusing human effort on the things humans are good at.