MBA / MS / Grad School · Primly Community

MS CS job search after graduation 2026: honest experience, took me longer than I expected

ml_mike · 6 replies

Finished my MS CS in December 2025, started searching in October thinking I'd have offers by January. It's now late March and I just signed. Sharing the honest version because the LinkedIn posts are all 'excited to announce' with zero information about the actual process.

Background: undergrad CS from a decent state school, thesis-track MS from a top-20 program, focus in ML. 2 internships during the program, one at a mid-size startup, one at a larger company in ML infra. Not going to name the school because it doesn't really matter for what I'm sharing.

What took longer than expected:

The ML-specific job market in 2026 is bifurcated. There's high demand for applied ML roles at companies with real ML infra, and there's basically a wall for pure research-adjacent roles unless you have publications or a Stanford/MIT/CMU pedigree. I had neither, so I adjusted and focused on MLE roles at companies with mature ML platforms.

The coding screen bar is still high. MS degree does not exempt you from leetcode-medium-to-hard. I failed two phone screens early because I hadn't kept my DSA sharp during thesis work. Spent 3 weeks doing nothing but problems before I started moving through loops cleanly.

Behavioral rounds are where MS candidates underestimate the gap. Most of my program friends had strong technical chops but zero STAR stories. The behavioral interview for a senior MLE role at a Series B or big tech is not softer just because you have a grad degree. If anything the bar for 'impact at scale' goes up.

Visa timing if you're international. About half my cohort is on F1/OPT. The cap-gap anxiety is real. Some companies are still skittish about H1B sponsorship even for strong candidates, so factor that into target list prioritization. Apply to companies with known sponsorship history first.

What worked: targeting companies where my thesis topic (recommendation systems) was directly relevant, doing explicit research on what each team was building before technical screens, and being honest in behavioral rounds about project scope rather than inflating it.

Final offer was MLE at a growth-stage company, Bay Area, good comp for a new grad MS. Not the FAANG dream but I'm genuinely excited about the work which I can't say I was expecting to feel.

Happy to answer questions. Especially about the MLE loop structure at mid-size companies if that's useful.

6 replies

ds_dmitri

The bifurcation point is accurate. I finished an MS in data science 18 months ago and watched the market shift even in that time. The 'data scientist' title is getting compressed: junior DS roles increasingly want ML eng skills, senior DS roles want some product sense. Pure statistics background without coding chops is struggling. Your advice to adjust the target function early is right.

ml_mike

Exactly. I had a few interviews where the JD said 'ML Engineer' but the actual work was closer to data analyst. And vice versa. Worth clarifying in the recruiter screen what the split actually looks like day to day.

visa_vik

The F1/OPT timeline thing deserves its own post honestly. You graduate December, OPT starts, you have a 90-day unemployment cap, and if you're in STEM OPT extension territory you need to be employed within a window or your status is at risk. I watched friends make suboptimal job decisions just because of timing pressure. Apply early and wide if you're international.

pivot_pat

The behavioral prep note is making me nervous because I've been 100% focused on leetcode and system design. How much time would you put into behavioral prep as a percentage of total interview prep time for an MLE role?

ml_mike

Honestly for MLE I'd say 30-35% behavioral if you're at a stage where the technical is solid. For senior roles or anything with a team-fit round it's higher. The technical part filters people out first but behavioral is what loses you the offer at the end. I failed one late-stage loop purely on a 'tell me about a conflict with a cross-functional partner' answer that I hadn't prepped.

market_realist

5 months for MS CS grad in 2026 sounds about right to me. I know people treating 3-4 months as a failure but that's not realistic anymore. The people announcing offers in 6 weeks are mostly people who converted internships or had warm intros. Cold application timelines are longer.