DIGITAL SOLUTIONS
Innovation & Digital Solutions Lead at Mana Battery. I turned the company's solar-and-storage know-how into web tools customers could use on their own, and built an AI system that automates content creation.
From sales conversation to software
Mana Battery sells solar and backup-power systems. Every sale used to start with the same engineering questions, answered by hand: how many panels, how much roof, how many batteries, and when does it pay for itself.
My job was to find those repeated questions and turn each one into a tool. I designed and developed the software myself, from the engineering model underneath to the interface a customer actually sees, and worked on the company's wider digital transformation by mapping operational needs to practical software.
- Solar sizing toolLocation and demand in, panel count and array area out. Or the reverse: area in, generation forecast and payback out.
- UPS configuratorRecommends and configures a battery package from each customer's load and backup requirement.
- AI content botAutomates image and text generation for the company's content.
- Digital transformationIdentified operational needs and turned them into software.
How many panels, and when does it pay back?
The tool works in both directions, because customers arrive with one of two questions.
In DEMAND mode you give a city and a daily energy need, and it sizes the array from the local peak-sun hours and a system performance ratio. In AREA mode you give the roof you have, and it forecasts what that roof can generate. Both feed the same techno-economic model: installed cost against yearly savings, with panel degradation, giving the payback year and the 25-year return. The curve is cumulative cash flow, and payback is where it crosses zero.
A simplified rebuild of the sizing logic for this page. Sun hours are approximate annual averages and prices are illustrative.
From the load to a battery bank
"I need backup for this load, for this long" is not a product. Turning it into one is an energy balance plus a set of practical rules.
The configurator takes the load and the backup time, chooses the DC bus voltage, and works back through inverter efficiency and the battery's usable depth of discharge to the amp-hours the bank must hold. Then it picks a standard 12 V block, wires enough in series to reach the bus voltage and enough strings in parallel to reach the capacity, and sizes the inverter with headroom. Switch to lithium and see the bank shrink, because more of each battery is usable.
A bot that writes and illustrates
Content was a steady, repetitive cost: every product and every campaign needed both text and images.
I designed and developed an AI-powered bot that automates both. Most of the engineering is in the prompts: structuring what the model is given so the output stays on brand, technically correct about the products, and consistent from one piece to the next, so a person reviews it instead of writing it from zero.
- BriefProduct, audience and purpose of the piece.
- TextStructured prompts generate the copy in the brand's voice.
- ImageThe same brief drives an image-generation prompt.
- ReviewA person checks and publishes, instead of starting from a blank page.
Engineering models, shipped as products
Each tool had to be right as engineering and simple as a product, for customers and colleagues who would never see the model underneath.
// HARD SKILLS
- Solar PV sizingPeak sun hours, performance ratio, array sizing and generation forecasts.
- Techno-economic analysisPayback, ROI and cash-flow modelling with degradation.
- Battery & UPS sizingEnergy balance, depth of discharge, series and parallel configuration.
- Web application developmentJavaScript front ends that wrap the engineering models.
- Artificial intelligence · prompt engineeringStructured prompts for reliable text and image generation.
// SOFT SKILLS
- Requirements gatheringTurning a sales conversation into a precise specification.
- Product thinkingDeciding what the user sees and what stays hidden.
- Cross-department workSales, technical and marketing teams feeding one tool.
- Explaining resultsMaking a payback figure or a battery bank understandable to a customer.
- OwnershipFrom the first question to a tool in daily use.
Engineering, packaged as a tool
Sizing a solar array or a battery bank is engineering a customer should not have to wait for. Writing the model once, correctly, and putting it behind a clean interface made that engineering available to everyone who asked.