The strongest demand we’re seeing in tech is concentrated in more advanced and specialised technology fields, particularly among professionals looking to future-proof their careers and expand their impact.
Many in our community have joined our courses for those exact reasons which includes Conor.
Underneath, he explains how developing the right skills can help you move from using technology to creating powerful AI and automation solutions that drive real business impact
Watch Conor’s interview to learn why a Google Solutions Architect with nearly 10 years of experience chose to invest in his future by attending Le Wagon’s Data Science & AI Bootcamp.
He explains:
Full Interview transcript
Jamie Speight — Interviewer
Conor — Alumnus and Solutions Architect at Google
Jamie (00:03): Hi Conor. Thanks for joining us, especially on a busy Friday.
I’m creating a video series about real people doing real jobs in tech. Students and people I coach often say it’s hard to get an honest understanding of what these jobs look like.
You’ve kindly agreed to walk us through your work and share your thoughts on the market and where AI is going.
To start, where are you working, what’s your role and what does your day-to-day work look like?
Conor (00:52): I work at Google on the advertising side of the business, in an organisation called gTech. You could describe it as a sales support organisation.
Google has people selling advertising across properties such as YouTube and Search. gTech helps advertisers, agencies and partners solve technical problems with those advertising platforms.
We also help them explore opportunities to combine the advertising platforms with technologies such as Google Cloud Platform, Gemini and ADK.
We bridge those technologies to solve business problems, usually marketing problems. That might mean helping a company grow sales through Google advertising or advertise more efficiently.
Jamie (02:21): Do you spend much time with clients, or is your solutions architect role more focused on strategy and building solutions?
Conor (02:28): Many people on my team work directly with clients, but I’m assigned to agencies and partners.
A lot of our big Australian clients work with media agencies or technology partners. I help those organisations use Google’s marketing technology on behalf of their clients.
Jamie (02:59): You’ve been at Google for a while and completed a course at Le Wagon. Which course did you take, and what motivated you to do it when you already had experience in tech?
Conor (03:21): I took the data science course when I was 31, about two years ago.
Before coming to Google, I already had eight or nine years of experience across media agencies and startups.
I started by working directly in advertising platforms and picked up technical skills along the way. At startups, I worked in cross-functional teams with engineers and data scientists and began learning SQL.
At one company, we needed to query tens of millions of rows relating to Google Ads keywords.
Working with engineers and data scientists, using SQL and seeing advertising workflows become increasingly automated all contributed to my decision. That automation was already happening in 2016, 2017 and 2018.
I wanted to move from analysing data with some technical skills to building and training models, or using existing models to solve more complicated business and analytical problems.
I also had friends who had completed Le Wagon’s software development course. They gave me a ready-made example and helped inspire me to do it.
Jamie (05:40): What problem were you trying to solve for yourself? What was the main driving factor?
Conor (05:52): Most of the technical skills I picked up before the course came from working at GetYourGuide, a travel technology company in Germany.
I worked in a cross-functional marketing team. Many of our business problems involved automating paid search, paid social and display advertising campaigns across platforms such as Google and Meta.
For example, how do you determine the right bids for Google paid search campaigns at scale? You need a model, but how do you build it, choose the right one and test whether it’s configured properly and better than an off-the-shelf option?
Other problems involved managing tens of millions of keywords or a catalogue of thousands of tourist activities. How do you manage a feed for that catalogue on a platform such as Meta?
When you have so many items to manage, how do you avoid needing lots of people manually uploading things to advertising platforms?
There were implementation and marketing automation questions, as well as data science questions about how to approach models for tasks such as automated bidding.
Jamie (07:43): You wanted to automate work and become more effective and efficient.
Before you became more technical, what did you imagine working in tech would be like?
Conor (08:06): I didn’t have many preconceptions. I fell into it at GetYourGuide.
When I moved from a tech startup to a big tech company, I expected everything at Google to be sorted and highly automated.
The reality is that in an organisation that large, different parts operate differently.
I was surprised by how much opportunity there still was to automate operations, particularly on the advertising side, and use our own technology to streamline how we work.
Jamie (09:23): When did you realise it wasn’t just about tools and code, but about your approach, your thinking and the skills and experience you already had?
Conor (09:50): When I started learning technical skills, beginning with SQL, I viewed coding as one single thing.
I didn’t appreciate how different coding for data science and analysis was from software development. I only understood that later, when I tried both to work out what I wanted to focus on.
I also took free online coding courses. They’re good for learning syntax, but they don’t necessarily teach you how to think.
Learning syntax isn’t the hard part. The hard part is thinking in the way required to develop a product.
You need to understand the objective and assumptions, then structure the logic so a computer can understand it. That’s very different from communicating with another person. There’s no room for assumed context; you have to be precise.
That may be changing slightly with LLMs and agents involved in software development, but the difficult questions remain.
In data science, does your model make sense given the assumptions and limitations of your data?
In software development, how do you structure the logic so it works correctly and achieves the intended objective?
Anyone can copy some syntax or download a model. The harder part is understanding whether it’s right for your problem.
Jamie (12:31): That’s one of the challenges with AI: people are using tools or creating things without fully understanding them.
“Solutions architect” sounds mysterious to someone like me who isn’t technical. Could you explain the role simply?
Conor (13:04): Job titles vary across the tech industry, so a solutions architect can mean different things in different companies.
In gTech at Google, we’re essentially technology consultants.
Google has several major advertising platforms, alongside Google Cloud Platform and its many services. There’s also Gemini, which appears in different forms across Workspace, advertising platforms and Google Cloud.
That amount of technology creates complexity for advertisers, agencies and partners.
Our role is to have solid technical fundamentals, understand a wide range of problems and investigate different Google products to find solutions.
We’re generally closer to generalists than specialists.
In practice, that involves consulting. An agency might see two or three ways to use Google technology to improve its advertising. We help it understand which approach is best.
During implementation, it might encounter errors or blockers. We help resolve those too.
The core of the role is understanding the business problem and finding the right marketing solution using Google technology.
Jamie (16:07): It sounds like a combination of consulting, understanding the end user and knowing how to build or work with developers.
You need to see the whole picture while also being able to investigate the details.
If you had to simplify it further, what’s the main problem you’re paid to solve?
Conor (17:28): Most of the problems are marketing problems for advertisers. The ultimate aim is to improve marketing efficiency or drive growth.
Simply put, how should a company use its marketing investment to grow its business?
The details depend on the business, what it sells and what it wants to grow, but it generally comes back to that.
Jamie (18:06): Your role is to take that complexity and turn it into something aligned with the client’s needs and the problem they’re trying to solve.
Looking back, which skills or experiences have mattered more than you expected?
Conor (18:34): Do you mean skills or experiences?
Jamie (18:39): Either, particularly ones you brought with you.
Conor (18:44): Being forced to get my hands dirty and use technology to solve real problems has been valuable.
You can take all the free online courses in the world, but you have to put the theory into practice. You need to repeat that process to build experience alongside your theoretical knowledge.
The skills I value most are a motivation and desire to keep learning. None of this technology is static, and it’s changing rapidly.
You also need the grit to see the learning process through. Learning isn’t a smooth, linear progression. Some concepts are difficult, and you have to stick with them.
First-principles thinking and strong analytical skills also help. I don’t necessarily mean working with numbers; I mean being able to analyse and synthesise information.
But motivation and persistence are the most important ingredients. You can learn almost anything online, often for free, but without those two qualities, you won’t take advantage of it.
Jamie (21:00): When someone joins your team, what makes you think, “This person gets it. They understand why they’re here and what they need to do”?
Conor (21:18): Google is a large organisation, and any large organisation has plenty of potential distractions.
Communication becomes more complicated when thousands of people need to communicate with one another.
When someone joins my team, I think they get it if they can minimise that noise and focus on the core purpose of the role.
For us, that means solving business problems for clients, advertisers and agencies, usually related to marketing.
If they keep those organisations’ interests paramount and prioritise that work over the distractions, they understand the role.
Jamie (22:32): You’re describing someone who understands how their skills and experience relate to the role and can apply them alongside new knowledge.
Moving to future-proofing: what should people focus on now to prepare themselves for the future?
Conor (23:16): There’s a lot of hype and speculation in the market.
We have powerful new technology, with LLMs driving a major shift from around early 2023. But it’s impossible to know exactly how it will develop. There are many possible outcomes, and plenty of commentary at the extremes.
Putting that aside, think about what probably won’t change. Businesses will still be trying to sell goods and services.
People who learn faster, enjoy learning and dedicate time to it will have an advantage. I also believe people with technical skills will continue to have a significant advantage.
One division in the tech community is between people who believe LLMs and coding agents remove the need to learn coding fundamentals, and people who believe those skills still let you do much more.
I’m in the second group.
LLMs help people with little or no coding experience develop products, and that will probably become increasingly true.
But if you’re vibe coding entirely through prompts, without coding experience, you’re limited in the complexity of what you can build and your understanding of it.
You may create good solutions, but can you investigate the code, understand what the LLM wrote and compare it with what you wanted?
We’re in a paradoxical situation: we won’t need to write as much code as before, but having the fundamentals—the ability to read, write and modify code—still gives you an advantage.
There’s a spectrum of vibe coding. Those skills let you do it more effectively.
Jamie (27:02): That’s a valuable point. I work with entrepreneurs who can already code and build, and they’ve never been able to ship products so quickly.
There’s an opportunity to solve real problems more cheaply and with smaller teams, especially if people can bring several skills together.
What does someone who uses AI well do differently from someone who doesn’t?
Conor (28:14): This form of AI has only been around for a few years. For some problems, it has only become particularly good in the past six to twelve months.
There’s still a lot we need to learn through repeated use and shared experience.
Someone who uses AI well can be more productive. As you use it, you develop an understanding of its limitations and where it works well.
For example, LLMs can help automate relatively abstract problems in ways that weren’t previously possible.
But for some tasks, such as complicated mathematics, my experience suggests traditional software approaches may be more appropriate.
Whether you’re developing with AI or using an AI tool, I think it will generally make you more productive.
That may not apply to every individual task today, but I’m confident it will increasingly be the case. If you don’t start using this technology, you’ll eventually fall behind.
Jamie (30:23): I’ve seen that in my own career coaching business. I can create detailed summaries that give people something tangible to take away, and build a workshop in a day that previously took two weeks.
For someone without technical experience, perhaps still working in marketing, what should they focus on over the next six months to a year to position themselves well?
Conor (31:28): Not knowing the future with certainty doesn’t mean you shouldn’t prepare. There are better and worse ways to prepare for the different futures we might face.
First, ask yourself how involved you want to be with this technology. I wouldn’t advise someone to learn coding—or anything else—if they aren’t interested in it.
Then work out what you’d like to do and create a learning objective. Make it specific and time-bound where possible.
That might be, “I want to learn these technical skills and complete this course by this date because I want to move into this role.”
It isn’t always that straightforward, so take small steps. I’m a big believer in incremental improvement.
You don’t need to become a software developer or data scientist in two days. That’s impossible, and it wouldn’t make sense to try.
Start using AI models in your day-to-day life, including personally if you aren’t using them at work.
When you start using them at work, identify where they’re most useful.
If you become interested in building or shaping things with AI, consider courses. That could mean free, self-paced online learning or a more structured bootcamp.
Conor (34:16): Think about how you learn best.
I learn best in a structured course with other people learning alongside me. That’s why I chose a bootcamp.
As you find a course and become clearer about what you want to do with the technology, the path becomes clearer.
Work out what you want to learn and why. Make sure you’re interested, keep learning, see things through and find practical ways to use your skills.
The learning never stops. New tools appear and people need to learn about them.
[A specific product reference here is unclear in the original transcript.]
Don’t overwhelm yourself by trying to reach the end point immediately. Start small, use AI, think about the technical skills you need and decide whether you’re interested in learning them.
Jamie (35:43): Three things stand out to me.
First, curiosity: get out there and experiment. It doesn’t have to lead directly to a job.
Second, skill stacking: making small improvements over time.
Third, learning and adapting as things change, then returning to that curiosity when something new appears.
To finish, what mindset does someone need to succeed, whether they’re moving into tech or using technology in another role?
Conor (36:53): You need curiosity and consistency in your learning.
Make time for it, whether that’s development time at work or time outside work.
Be an active learner, want to solve problems, show initiative and have the grit to see things through. It isn’t always straightforward.
Learning, hard work and consistency are the defining traits I’ve seen in people who learn well and pick up new technology quickly.
The reason that matters is that we don’t know how the future will unfold.
If you have strong fundamentals, can think logically and add technical skills, you can pivot in different directions.
If you also understand businesses and have experience solving problems in your existing field, you have overlapping areas of knowledge that can give you a unique advantage.
We don’t know exactly how this technology will affect society. But strong fundamentals and that learning mindset put you in a good position to adapt.
I’m convinced most people can learn most things. We now have more free information at our fingertips than ever, through technical blogs, YouTube and other resources. There are also bootcamps such as Le Wagon.
If you build a broad foundation around a problem you’re passionate about solving, you’ll be well placed to pivot as the technology develops.
Jamie (39:56): We’ll leave it there. I’m going to sign up for a bootcamp somewhere.
Jamie (40:03): The purpose of this series is to help people understand what working in tech looks like, and this has been an honest, practical perspective.
I’ve learned a lot, and I’m excited. I’m on the optimistic side, and that’s been an important part of this conversation.
You may have created more questions than answers for some people, but that’s part of curiosity and learning.
The hard work still matters. It isn’t only about taking a course. It’s about understanding where you want to go, what you need to do and having some fun with it.
Is there any final advice you’d like to add? I thought your answers to the last two questions were fantastic. Thank you for today.
Conor (41:05): No worries. I think that’s everything.
Jamie (41:08): Thank you for your time.
Conor (41:11): Thanks.
If I had to choose 3 words to describe Le Wagon, I would choose positivity,
What principles should guide AI in the workplace? Québec’s Ministère du Travail outlines five principles
After just two months in Le Wagon Montréal’s Data Analytics bootcamp, Coline walked away with
We often hear that we are in an “data-driven world.” And in an era fueled
In today’s data-driven world, businesses are recognizing the value of turning data into actionable insights.
Las call to Le Wagon 2021 Spring Bootcamps! Last Day to turn in your application
