Month: November 2024

Organizations don’t succeed because of the little things they do well; but they often fail if they don’t do such things well.. These are the minor hygiene tasks that though doing them well doesn’t necessarily do any good, but not doing them comes with significant repercussions; sometimes existential. Try skipping washing your hands for a week. You’d probably land in the hospital.

When the Bible was talking about the little foxes that spoiled the vine, trust me, it was about these tiny annoying stuff.

This becomes a catch-22 for companies, especially medium to large ones. How could they focus on what’s important without getting lost in tactical processes that don’t guarantee success even if we do them well?

Automation to the rescue!

I remember sending an email to my team a couple of years ago where I said that tactical work done with lack of automation can bring out the worst in the best of the best.

Speaking about myself, everybody knows I have the sh*ttiest, most incredibly stupid memory in the world. Half of the time when I wake up in the morning, I can’t remember my own name. I literally have a document by my bed that reminds me my name is Adedeji 🥹.

However, one of the ways I’ve solved my problem is that I figured out that I could use a digital crutch  to remind me of all my pending tasks. For this, Todoist has been pretty incredible.

This has helped me extremely as I use it to set reminders to check on my siblings, check on some elders who haven’t yet kicked the bucket, reach out to those seniors to me and even reach out to some customers and some of my employees. I know fully well that If I don’t set these reminders, I won’t be able to do these things. 

Basically, without some form of automation, I’d have been messed up.

I also noticed when I started Lendsqr that there are some incredibly great guys working with me, yet I get so upset with them because packets are getting dropped. The funniest thing is that they weren’t dropping the large, extremely important packets, but the much smaller ones like sending reports and checking system performance.  

That was when I told myself, you know what, any organization that doesn’t have a way of automating its tactical stuff will never go anywhere because they would keep dropping the “hygiene” tasks that are not so large but are highly important.

This brought me back to when I was at FCMB (I’ve worked in a million banks) and we implemented this back office automation application called Integrify. I worked on this with some amazing people and I’m always excited when I remember the bank is still using that till today. That was one of the best projects I ever worked on.

I tried to do this with every single bank or institution I ever worked in since that time like UBA and Fidelity Bank, but the cost  was always high, and no one really cared.

But I understood what it meant to be efficient, and so when I started Lendsqr, and saw the same problem, where the best people were getting dinged by customers and sometimes, me for dropping some packets, it just occurred to me that we couldn’t and wouldn’t win if there were no automations, especially for the tactical tasks.

So the question was, how would we do this?

There were a few things we worked on, such as running jobs in the database to send emails and tons of Metabase results, but that wasn’t enough for what I wanted to achieve.

We knew something had to be done.

We kept experimenting until one of us, Frogman, found n8n, an open source automation software, which we deployed and that has been one of the best things we’ve ever done.

n8n is like Zapier, though not as fancy and popular. It’s an open source software that helps you automate tasks, with plugins for almost every kind of task you wish to carry out. Fortunately, some smarty pant drank my Koolaid and joined my team so I nailed the automation initiative to her cross, to bear.

We then went on a tear and automated as many things that could move from the HR process, to Sales, Product Support, and Finance processes – It’s absolutely amazing. 

N8n was so much fun that when Bland AI was launched, we used n8n to automate our introductory calls to our customers. We also used it to remind some of the  loan defaulters at Irorun. Those usually swear for us and then never paid. Nigerian borrowers are something else 😪 

We also use it for Sales to notify my team instantly whenever a non-Nigerian lender signs so we can reach out immediately and engage on WhatsApp.

System Performance

One of the worst things that could happen to a company is having things breaking in the background and no one is aware of this. Not every customer would escalate to the team. You might just come back to realize you’ve been using your own hand to chase customers away.

At Lendsqr, we’ve had our fair share of this, finding out certain things are breaking and sending our customers away without our knowledge. While there are reports that can be sent constantly by team members, there’s just so much the team can report on. This is where n8n comes in again. We have automations that notify us when customers are complaining, when certain system metrics are lower than they should be. We automatically test our systems and report specifically on each of them to prevent stories that touch the heart.

These are just some of the examples of the use cases we have set up.

In summary, n8n is one of the best things we’ve ever done as a company and I’m looking to keep finding better ways to improve organization efficiency and provide the best experience for our customers.

As of this month, the general AI frenzy across the world would have been around for about two years. Although, in reality, AI has always been there, it’s just that the buzz around it started much later. It’s been almost two years since Open AI launched ChatGPT in November 2022, fueling a widespread popularity of AI and since then, AI technology has unveiled some mindblowing developments.

At first, I felt the trend, just like other useless problem-seeking-solutions, will fade. But it hasn’t so trust me to start experimenting. 

Being a pragmatic soul, I found some commonsense AI tools play around which include Bland AI for voice calls, Visual Electric for image generation, and Google’s ImageFX for podcasting. 

Bland AI as it allows you to have proper intelligent conversations with the tool but it still feels a little script kiddie and unpolished so I wouldn’t call it a favorite at this time; still a work-in-progress. Visual Electric was made for design and lets you generate so many images. It’s pretty amazing and this one is definitely a favorite. That’s what we currently use in my company, Lendsqr and even in Open Banking Nigeria,  instead of looking for stock images from Shutterstock or even open source images. Google ImageFX is also great and you can use that to create audio, images and maybe even videos soon. 

Do you see the big picture yet? 

In the last couple of years, AI technology has rapidly expanded to support a wide array of activities that before now, could take hours or even days of work to complete. In Lendsqr, we’re embracing this new reality and using AI to drive customer service and other activities where we’ve identified that integrating AI into the process can give us better results and free up our talent for more engaging work.

Bless AI …

As much as we can look at AI as the shiny new thing (for those who just encountered AI technology for the first time out of their television screens with ChatGPT), its rapid and widespread uptake is owed to its unmatched utility and ability to solve real problems that people like me, especially in Africa and other less developed regions, have been dealing with for years. 

The first pain is that being able to get the right type of image has always been a concern. At Lendsqr, we’re heavy on marketing because I love writing and I think it’s an effective way to get our message across to our target audience. Even on a personal level, I’ve been writing for about 23 years so I know the power of writing first hand. So, we do a lot of marketing, posting on social media, etc. and being able to get the right type of image for the markets we operate in, has always been a concern. This concern has grown most especially in recent years, where we have to be sensitive to demographic representation needs with our target audience.

The second is the high costs for image sources. We used to sign up for expensive tools like Shutterstock to access images we could use but it became clear that this came with the problem of a limited image stock. So even when you find good images, you’d realize that they’ve been used or will be used so often that if you ever meet the models by chance, say at the airport, you’d be so convinced that you know them because they’re on everyone’s billboards and social media posts. Before now, if you wanted something very creative for yourself or your brand, you’d have to shoot it. That’s also expensive and you’d probably only use the image once.

And for those who tried to escape the cost of these image sources and opted for free sources, 9.5/10 times, they were met with quality issues. Yes,there are some free sources available, but they’re not common and the quality is usually not great, or even decent enough to get by.

But AI technology came to save us from all that.

Here’s how AI technology has made my world a little brighter

Over the last year and some, we started seeing guys like DALL-E, Midjourney and Visual Electric come up with the ability to generate stunning and unique images. But many people can tell that those images are AI generated. However, they’re getting to a level where more people are getting confused, wondering if the images are AI generated or not. 

I watched this happen recently: my siblings and I were reminiscing about our mum’s old car, a blue Volkswagen Beetle, and in that wave of nostalgia, I got curious about how accurately I could reconstruct the key elements of those memories. So, I went ahead, dropped my prompt, generated an image of the car with a young lady standing by it and posted it on my WhatsApp status. Interestingly, people who know me reached out to ask if the lady was my cousin to which I responded “yes” and when pressed further to ask why they had never met her, I responded that it was because she was my AI cousin. Made for a good laugh. 

Although AI generated images may not always come out perfect, if you’re good at prompting, you can dictate what you want down to the smallest detail and master the greater creative control AI technology affords you. I envision that AI image generative tools will get so good that it might not even make sense to use stock images anymore because stock images are coming from somebody’s preconceived idea of what another person could use it for, while with AI, you get to generate images based on your specific needs.

Additionally, it definitely doesn’t hurt that AI tools are more affordable and give you more value for money. For instance, Visual Electric is dead cheap at just $16 a month for the standard version and you can even use the free version if your needs are limited or you just want to give it a try before a financial commitment. These tools create more images than you could get from stock images, and what you get is unique to you. 

Above all, I think the greatest attribute of AI technology is its ability to evolve. I was engaging with the CEO of Visual Electric recently, Colin Dunn, and he mentioned something about them working to release a version that’ll generate videos too. Amazing.

AI technology is impacting how we work in incredible ways 

At the rate we’re going, people will be able to generate tons of images that may even be better than real life. Soon, this gets to audio, music, videos, animations, and possibly, 3D. This way, people will be able to create interesting content more easily and much faster, and this obviously increases its utility for marketers.

Today, people snap photos and use filters. We know they’re using filters but nobody cares.

Just like that, a time will come when we’ll know that you’ve generated an AI image, and nobody will care—they’ll look interesting and normal.

In fact, your creativity would soon be measured from being able to use AI to generate extremely creative content.

There’s a lot of talk about AI replacing people but perhaps we also need to talk more about how it helps people become more efficient. In Lendsqr, we’re so crazy about hiring only the very best people; a lot of them are first class graduates, very highly talented people and when we assign them to do customer service jobs, they don’t like it. No brilliant person wants to go to school, just to end up doing customer service. And I’m strongly against the idea of lowering our standards to find those who maybe weren’t so great at school and want to do customer service because they think it’s easy. I want everyone who works in Lendsqr to be great and first class rated. 

So, we thought about it and instead of dumbing down the job requirements and bringing people that may not be fit for our highly intellectual environment, or making very smart people unhappy by doing mundane support jobs, we decided to increase the challenge and transform their jobs to high-level problem solving support. These guys are so smart that they’re now building tools to do the mundane jobs they don’t want to do. AI helps my people do great work.

Two-thirds of AI is a blessing and one-third is a wicked curse 

I know I’ve spent most of this piece talking about how great AI technology is but I won’t pretend to be blind to the adverse impacts this advancement will have on the way things are done, the people who do them and those who consume the output.

People are going to lose their jobs for sure, because they may not be that great, or even when they’re that great, they’re competing with something that costs less than $20 a month and can work 7 days a week, 24 hours a day. These tools are always available and they’re learning very quickly how to understand what people need.

Another unpleasant outcome is that customer support business process outsourcing (BPO) centers in India will become obsolete. And this would happen so fast that Africa may never have enough time to get a slice of this.

Also, while most of the benefits of the increased efficiency that comes with AI are clear, the overall impact on society and advancement of the human experience is still unclear. The technology is still developing, so there’s a big question mark on the quality of results and ethical concerns which can have adverse effects on the quality of decisions resulting from AI input. Can we trust AI generated data enough to make business, economic or life-altering decisions?

I admit that AI has been great for improving overall efficiency in my company but it can only go so far. There are some parts of our human experience that deserve the delicate human touch and original thinking that AI is yet to replicate. Can we objectively say that work devoid of that human touch and unique creativity is truly great? I’m not so sure. I’m a grade A tech fanatic but even I don’t leave certain things to AI. 

Regardless, the benefits outweigh the costs

While it’s true that some people will lose their jobs to this tech advancement, the beautiful thing is that generally, we’re going to be better off. The progress made with AI technology means that a lot of people today, for whom being able to have armies of people to provide support, is what has been the cap on their aspirations of building very successful global businesses, can now do so with little to no friction.

The 24/7 availability of these tools also means that businesses can expand their internal functions faster and cheaper. Even running even a one-man business is a lot easier now, with tools that can do customer support, marketing, accounting, etc.

You can also train AI faster and without incurring thousands of dollars paying for employee training courses. These tools will learn and grow much quicker than human beings which means your business and your quality of delivery can improve in much less time—your 5-year profit projections might very well be within your grasp in less than half of that time. And they’ll enable you to more easily eliminate human error to produce more consistent results over time. 

I  believe I’ve just painted the ideal growth conditions for African entrepreneurs like me. 

So, what’s holding you back?