My one year of AI in Reaktor
I recently reached my 1-year mark at Reaktor as a software consultant and, when looking back, I was surprised at how much stuff had happened in that year. For example, I got four AWS certifications and a Scrum Master certification during that year. And it has also been a year since I started journaling seriously every week in Obsidian. Most outstanding, probably, has been my focus on AI during this year.
Even though AI has been hyped since ChatGPT's public release back in 2022, I had mostly been by the sidelines. Back in university I took a computer linguistics module that I felt passionate about, and I also spent 2 years at Samsung (2018-2020) localizing their voice assistant (Bixby) to German. When I tried ChatGPT for the first time I realized my previous AI experience had been quite primitive.
When I joined Reaktor, I didn't have a project lined up immediately. So I decided to finally dive deep into AI and see by myself what all the hype was about. AWS had just released two brand new AI certifications, and I decided to tackle AI Practitioner (AIF-C01), studying while I waited for a gig. I knocked out AIF within 3 weeks, and then took the next certification, AWS Machine Learning Engineer - Associate (MLA-C01) some months later, while working on my first gig. Preparing for these 2 certifications got me up to date with AI: I learned about traditional Machine Learning, Deep Learning neural networks, and GenAI (especially transformers and LLMs). The knowledge was high level, but enough to complement my skills as a software engineer.
During this year I had two client gigs at Reaktor, and both had mainly to do with an AI chatbot's backend. The AI knowledge gained from the fresh certifications was very helpful to understand stuff like RAG and vector embeddings right from the get-go. Both gigs were short (3 months each), but I did gain important experience with AI projects that are becoming very popular in the software industry right now.
Apart from my client gigs, I have been very active working on GenAI within and outside Reaktor. I have been part of Reaktor's internal GenAI team to spread awareness and training to the whole company. I have also organized meetups and events internally and externally (e.g. collaborated with Symposium AI to organize an AI event 2 weeks ago at Reaktor's premises).
Was it all worth it? I do think so, as adding "AI" to my job title in LinkedIn has made me much more prominent. I used to show "Software Engineer at Reaktor" when I started a year ago. Some months later I updated that to "AI Software Engineer at Reaktor" and I've gotten more interactions, more connections, and more job offers. I do think GenAI is overhyped, but it's also a reality that it's a really hot topic right now and that software engineers who upskill in it right now will see more opportunities.
One year of AI at Reaktor—Summary and Links
- Upskilled in AI via AWS certifications
- Two client gigs on AI projects (RAG chatbots)
- Member of Reaktor's GenAI team
- Wrote a summary of the HARVEST event in Reaktor's official blog
- Trained dev teams on AI developer tools for University of Helsinki's research in collaboration with Reaktor
- Organized internal and external AI events