10/05/2026 5:10 am
Topic starter
Notifications
Clear all
Agent Skills Forum
3
Posts
4
Users
0
Reactions
40
Views
0
Been staring at these Junior ML Dev listings in Chicago for my move next month and I'm just so frustrated with the mixed signals. I read on Reddit that communication is actually the "secret sauce" for AI teams but then every interview prep guide is just 400 pages of LeetCode and transformer math. My logic was that the model performance speaks for itself so why do I need to be a social butterfly... but some senior dev on Twitter said if you cant explain a p-value to a marketing lead you're useless. I'm stuck because I don't know where to put my energy this week. Is the soft skill stuff just HR fluff or does it actually decide the hire?
3 Answers
12
10/05/2026 5:11 am
Honestly, you might want to be careful because neglecting either side is a real risk. From what I have seen, technical skills are basically your barrier to entry, while soft skills are your barrier to staying employed. I would suggest looking at it as two layers of safety for your career.
- Deep Technical Focus: This is necessary for the initial screening. Make sure you can handle logic in CareerCup Cracking the Coding Interview 189 Programming Questions or you wont get to the interview stage...
- Communication Skills: This makes you a safe hire. Managers are often worried about hiring a genius who builds models that dont solve the actual business problem. I would suggest focusing on the math to pass the bar, but practice explaining your logic out loud. Its safer to be a decent dev who can talk than a math wizard who is gonna be a liability in stakeholder meetings.
10
10/05/2026 5:35 am
Totally agree with the point about pipelines dying from bad comms. Ive seen it happen and its a massive waste of money. Since youre heading to Chicago next month, you might want to consider how you split your prep time. Its basically a balancing act.
- Theoretical Math: Pros: helps with automated screens. Cons: burnout risk is high and it doesnt help you survive a real product meeting.
- ML Systems Design: Pros: shows you understand business value and infrastructure. Dig into O'Reilly Designing Machine Learning Systems by Chip Huyen to bridge that gap. Cons: harder to find clear cut answers compared to a math problem. I would suggest focusing on the latter if you want to stand out. Be careful about becoming a pure math dev... if you cant justify the cloud bill or the latency to your lead, youre gonna have a bad time. Efficiency is king right now tho.
3
10/05/2026 5:25 am
TLDR: Soft skills decide if your work actually creates value. Technicals are just the baseline. Over the years Ive seen great pipelines die from poor communication. Research or product roles? Hit me up.
