How to get into AI as a software engineer
What you need to know. How to build up your profile. How to prepare for interviews..
Hey reader, welcome to the 💥 free edition 💥 of my weekly newsletter. I write about building startups and careers in AI. You can learn more about me here. Feel free to send me your questions and I’m happy to offer my thoughts. Subscribe to this newsletter to receive it in your inbox every week.
AI is the goal and ML is a vehicle to get there. Many software engineers are aiming to transition into the AI field. If you want to get into it, you should start by conducting your research on what you want to do within the field. I've published a post on that topic here.
You should talk to AI professionals at various companies. If you don't know anyone, just find relevant people on LinkedIn and ping them. You can tell them that you're looking for guidance. Many people are responsive and are happy to spend 15 mins with you on the phone. Always be hustlin'.
Here are a few companies that are always hiring for AI roles:
Google, Netflix, Meta, Amazon, Apple, Nvidia, Microsoft, Snowflake, Stripe, OpenAI, and many AI unicorns.
To prepare yourself to get a job in AI, I've outlined 10 steps that you can follow:
1. Decide what you want to do
You need to develop skills to slice and dice data. You need to decide what type of data interests you:
numbers, images, video, text, sensors, audio
I've published a post on it. You need to make a decision on what type of data you want to work with and develop expertise in that area. AI is becoming ubiquitous, so you need to have a point of view on what you want to do.
2. Decide the sector that interests you
AI is used across many sectors such as:
Fintech, Gaming, Ecommerce, Healthcare, Security, Physical Infrastructure, Robotics, Manufacturing, Media, Autonomous Cars.
You need to decide what sector interests you and make sure to have good knowledge about that sector.
3. Learn the tools
You should be familiar with python, tensorflow, ML libraries, command line, jupyter, cloud deployment tools, automation tools, and anything to do with deployment.
You need to know what tools are used to build real world AI applications.
Many people only have experience doing toy projects with clean data. Real world is messy. Your value goes way up if you know the tools needed to make things work in the real world.
4. Learn statistics
You should know statistics terms such as mean, median, gaussian, standard deviation, and other related concepts. Many roles in AI require you to look at a real world problem and break it down into various components.
There are no certainties in the real world. Only probabilities.
The sooner you can train yourself to think in terms of probabilities, the better off you'll be. Be well versed in statistics.
5. Learn how to work with data efficiently
Data will be an integral part of your existence as an AI professional. Make sure you know how to work with data efficiently.
80% of your time will be spent on organizing the data
You should know how to automate data-related tasks. This includes performing operations on data such as parsing, storing, retrieving, and moving. If you don’t know your way around it, it’s going to be a long slog. You’ll get tired quickly.
6. Learn algorithms and frameworks
You need to know what algorithms and frameworks to use in any given scenario.
Is this a classification problem or a regression problem? Do I need to use supervised learning or unsupervised learning? Should I use Random Forests or Hidden Markov Models?
Be familiar with the landscape and know what's applicable where.
It comes with practice, so practice a lot.
7. Build applications
All the learning will be pointless if you don't know how to build. You should know how to make AI work for an actual use case.
You need to build applications. Implement projects and get yourself familiar with operationalizing AI.
8. Write about it
Writing is a good skill to develop. No matter what you do, you’ll get better at it if you can write well. More than anything, it will help you clarify your own understanding of AI concepts. It will fill up any gaps in your understanding.
Writing can be quite cathartic
In addition to that, it will help others who are in similar situations as well.
9. Reach out to recruiters
If you're looking for a job, you should reach out to recruiters and work with them. It usually works much better than applying online on the website and waiting for a callback. You need to hustle your way in.
10. During the interview
Show initiative during the interview. At the end of the day, companies want people who can build. It's up to you to prove it to the interviewers. Show that you're willing to go the extra mile. Companies are moving away from the old "puzzle solving" interview sessions.
Puzzle solving doesn't tell us anything about the interviewee's building capabilities
Companies are moving towards the take-home assignment approach. Be prepared for that.
Where to go from here
If you know how to write production-ready software, you already have a leg up on a large number of people who want to build a career in AI. You just need to pick an area within AI and specialize in it. Many people just want to do everything. Why limit yourself, right? They're the ones who end up doing nothing in the end. You need focus and discipline if you want to build a career you want.