This page is a collection of various ideas a.k.a. “things I would build if if only I had the time”.
If you’re interested in building something from this list and want to talk about or even work together, get in touch.
Start with a responsive, mobile-ready spaced repetition flashcard app for the web.
Word lists are automatically compiled for any given language from Wikipedia and similar multi-lingual web resources. Foreign-language only flashcards (using only the foreign word and a representative image) are magically created using a combination of image search, natural language processing, smart heuristics, user feedback and other such cleverness.
The app will support a few ways of learning a lot of vocabulary fast, including alternative learning methods such as goldlisting.
It’ll be like Anki without the hassle of building decks, and Memrise without the reliance on user engagement or pre-constructed courses.
The site will analyse a bunch of popular websites for each language and will use word frequency analysis to figure out how hard to read they might be. Then as your vocabulary grows, it will suggest new websites or media outlets that you’ll now be able to understand without too much difficulty.
You’ll be amazed at how much you can read. And the more you read, the faster you learn the language.
Language learning is a huge industry, and an awesome application like this could really earn someone a living.
Figuring out how to get somewhere can be tedious. The fact that an office is 30 miles from you has no bearing on how long it will take you to get there. It could take 30 minutes if you have a motorway near your front door and traffic is good. But what about when the route isn’t so simple?
If you have to factor in tiny roads and traffic, or a bus trip to the train station and a short walk at the other end, or a cycle ride to the bus station followed by a tube journey, things get complicated pretty quickly. Not everyone drives, and not everyone would choose to drive over taking a short train journey.
Enter your address. Tell us your preferred modes of transport. Specify how much longer you’d be happy to sit in a car vs. on a train (or vice versa). Figure out how long you like to spend getting ready in the morning, if your arrival time is early.
The site will do the hard work of working out the best way for you to travel and even tell you what time to set your wake-up alarm. Traffic stats, local transport and your aversion to walking (or preference to get some exercise) are all taken into account.
Maybe I’m just scratching my own itch, but I feel like I spent 10 minutes going through this process every damn time I have a morning meeting on a client site. This stuff seems like it is too trivial to waste my time on. Let’s automate it.
See above. Take what you’ve built, and add job searching (with accurate office locations). Job hunters can search for roles within a sensible commute, and let the website do the rest.
In short, the site should automate the process of figuring out which jobs are worth getting to and which commutes will be unsatisfactory. “Within 30 miles” is a useless metric. Let’s kill it.
Cold-calling is hard work. Doubly so when you’re nervous or new to it. Yet there are a ton of small businesses in the UK, many of whom rely on cold-calling to drum up their business. They don’t need to be distracted by figuring out which steps they should go through on each call, recording who picked up and who was engaged, noting down the same action (or failure) over and over again, then remembering to follow up.
This stuff is fully automatable. The site should provide a basic workflow. Users enter their call list. Each one is dialed automatically, and the key steps are displayed on the screen to remind the caller what to say and what their next step should be. After each call they answer a short questionaire:
Short, simple, quick. And after a 20 second rest (with a countdown), the system starts dialing the next number on the list. No sitting around faffing between calls. No building up courage. Cold-calling is a numbers game – so let’s act like it.
You can choose to run a session by duration (60 minutes of calling), time speaking (10 minutes talking to clients), or call attempts (50 numbers to dial). At the end of the session, you get a nice breakdown from the site showing you what you achieved with pretty graphs, stats and follow-up todo lists (with reminders). Analytics are king, actions are key.
Any easily distractable business person would get a ton of benefit from this, in terms of saved time if nothing else. It scales to small teams, too. Plenty of opportunities for recurring revenue.
Hobbyist groups and hackerspaces are all about innovative homebrew alternatives to big industrial kit. 3D printers are all the rage, biohacking is increasing demand for cheap alternatives to gene sequencing and manipulation, and robot wars are as popular as ever.
But what about the poor materials scientists? EPR, NMR, xray diffraction, and electron microscopy are still pursuits for the lab. Most of the equipment needed for this costs tens of thousands of pounds (or more) and is simply out of reach.
Perhaps it doesn’t need to be this way. Components are now more readily available than ever, hackerspaces are appearing with well-equipped great workshops, and people are making homebrew tunneling electron microscopes in their garage.
Let’s apply the Makerbot business model to the wonderful world of materials science. Imagine open source equipment design with options to buy pre-packaged component sets or even pre-built spectrometers for those who want it.
Market kits towards undergraduate labs and hackerspaces. Find cool applications for small businesses and forge your own niches.