The Apply AI podcast explores how Artificial Intelligence and Machine Learning can revolutionise businesses. Solutions that save time and money, increase revenue and future proof their business.

Human Pixel: AI Solutions Putting Humans At The Centre Of Technology

Human Pixel is a software development company, putting technology to work to automate and find efficiencies for businesses. They put themselves in their customer shoes, setting realistic objectives and have challenging conversations in order to really put the ‘Human’ in AI and Machine Learning solutions.

Listen below or on your favorite listening service: YouTube | Spotify | Apple

The following summary was generated by Sonnant‘s SEO ready content & keyword transcription. Sign up for a free trial.

Free Sign Up

What is Human Pixel?

Human Pixel is a software development effectively at its core. It came together several years ago when my partner adam and I sat down and thought, what can we do that’s a little bit different? We’ve both been in technology for many years. We wanted to build stuff. We want to be innovative, creative, but we wanted to do it differently. We wanted to definitely make a difference. Really try and follow through on authenticity. We hear so many horror stories of people who go to get stuff built and it never turns out to be what they thought it was going to be. It’s often painful. And I think a lot of that comes from the mismatch or the disconnect between the supplier and the customer. So we set out to try and be real with our customers and set real objectives and also challenge our customers and say, “I hear what you’re saying. This is what you’d like. But have you thought about this?”

We want to be an adviser to them and not just a supplier of services. And that means putting ourselves in their shoes. There’s lots of pixel pixel companies, right? So the Human in Human Pixel really means putting ourselves in a customer’s shoes and really understanding what’s important to them and being able to deliver AI solutions

A lot of people think if you build my CRM for example, all of a sudden my life is different. We have to explain that this is part of a process. This is a tool, but you also may need to make some changes in the organisation, culturally.

Tell me your approach to software development.

We know people will often come to us with a project and I think that that’s it. However, there’s a whole planning segment that has to happen before that this. We often get clients coming to us saying , “I need this and so can you build it on this?” And we’ll hear that. And if we were just order takers, we’d probably just go ahead and build it. But that might not be the best way to go forward.

It’s about understanding the customer. if they have the knowledge then fine don’t challenge them. But if they think they have the knowledge and oftentimes we have clients come to us who think they know what they need. It’s really about exploring that out, pulling it apart and saying, well what is the real problem here? Don’t come to me with the solution. Come to with the problems. Let us provide a solution for you and if it matches up with what you think the solution is great, and if not, let’s explore what might be better possibly.

I think we always try and bring it back to the problem. What are you really trying to solve here? Because we can build the perfect solution to the wrong problem. And that can just be a real costly affair.

How hard it is for the different segments to apply software to solve problems?

It’s definitely not a one size fits all. And again, it’s about understanding what each individual business needs. Software on its own isn’t a magical fix. So if we understand what the problem is, let’s say it’s a CRM. We’re really trying to understand what a business needs in its sales, lead to sales conversion, ongoing support and follow-ups. So I often tell them don’t even think about the technology, tell me what you need to have happen in the business. When I understand that, then we can make a recommendations. What’s the best platform? What’s the best way to achieve it?

Often people come to us with a ready-made kind of idea of what they need. And it’s about untangling that and just saying, tell me the problem. Tell me what you need to have happened. Do you need more sales, more leads, more conversions? Tell us, and then let’s design the solution for that.

I pride ourselves on the fact that we’re fairly agnostic when it comes to technology stacks. So it’s not about trying to push something because that’s the window we represent. It’s really about understanding the problem, doing the real consultation. We’ll often start off with a discovery workshop and it’s all about just understanding the client problem. Then mapping out what that might look like in terms of a process that’s required. Then we take that away and formulate what the solution looks like. So we never talk about the solution first up. It’s about the what’s needed first.

Your approach to AI – what does that look like for the uninitiated?

For a small business might mean a chatbot on a website or facebook page. For a larger businesses, it might be about machine learning and large datasets and being able to create algorithms from those. So really all we’re saying is that we’re trying to automate logic.

So small businesses can set up a messenger bot on their Facebook page and have some very basic things automated – bookings, address, open hours. Very simple kind of stuff and that you can do fairly cheaply. If you want something further, which is not just a multiple-choice kind of bot, where it is free text, then we need to start talking about the different ways that the same question can be phrased. So there’s a lot more depth involved there.

We’ve built our own chatbots and effectively some of our clients have enabled the chat bot to be able to field inquiries outside of normal business hours. It’s there to assist during business hours and reduce the amount of content coming through to the contact centre. But it also fields half as many calls after hours, which were not going anywhere normally. So in terms of efficiencies & productivity – to be able to have a higher conversion rate, that’s when it’s coming into its own for some of those businesses.

So if you never had that chatbot, you would never have the chance to win some customers.

Correct. And depending on how advanced the chatbot is, the bot can actually handle the sales process. It’s not about – just leave your details and contact us – it’s about let’s facilitate this. Let’s show you information that’s likely to convert you. Once we are at this point, then a lot of businesses want to do automation as much as possible. The same reason that they don’t put phone numbers on websites anymore, if you’re going to go online purely. A way for people to engage at any time of the day. The web is 24/7 and people now like to sit in their bed at night on a laptop or an iPad and just do their research or book a trip or something. Now it’s expensive for a small business in particular to have staff there just to answer calls. So being able to automate that process and then go deep into your systems to be able to retrieve booking details via APIs. Then it becomes really useful.

A larger business may have more data & complexity. What does that look like them?

So particularly from a sales or marketing point of view, you want to be able to pull data from multiple touch points, aggregate that data and then be able to apply some logic and algorithms to it to then make some assumptions and then take further actions. So you really automating the process of not just pulling in the data, not just reading the data, but then taking further steps that either provide support or close a sale.

Usually some human intervention is done at the beginning to design what its going to look like. Then the heavy lifting is done by AI.

Where does the Machine Learning come in?

If I was trying to explain Machine Learning to someone, I’d say, if you’ve ever bought a brand new modern car these days, they’re impossible to understand. But what they do is they learn your driving behaviour and based on the way you drive that car in the first few days, weeks, months, it then sets up a style of driving for the future. It may run more economically by taking data from a driver. It keeps learning and changing based on those data set. That is a very simplified way of understanding machine learning.

How do you measure the return on investment?

So it’s about understanding what is the problem. Then we can work out what is the cost to solve the problem. And then we can determine what the ROI will be.

It’s not always the cost that’s important. I maybe reduce your workforce. You might be able to transition people into other higher-value tasks and have AI and automation implemented.

When it’s more of a qualitative thing. You might get better customer satisfaction. You might have more improved staff morale because the system’s easier to use. That can be harder to measure.

I think it’s really important that you are able to measure whatever it is that gets implemented. Because otherwise, you’re shooting in the dark if you don’t. And obviously, more sales is something that everyone wants to achieve. And that can also be difficult to quantify.

Other references that you can look at who has done a similar thing, so you try and find those references that will support the argument. But at the end of the day, every business is different.

What is your ideal business? Your sweet spot in providing a demonstrable improvement for customers?

We can add value to a small business very easily. We can do the same for a larger business, it’s just on a much bigger scale. And the solution may be quite different. It’s really about finding businesses where we can understand the problem. I keep getting back to that because if you can’t define the problem, then the solution may be hit or miss.

2021-07-23T02:02:15+00:00June 15th, 2021|
Go to Top