Uber Recruiter on CS Grads and the Tech Job Market in 2026

Introduction

The 2026 tech market is the most pessimism-soaked it’s been since 2008. CS grads on Reddit are posting application counts that look like phone numbers and concluding that the field is finished. That’s the doom narrative. What does someone who actually hires developers say?

A senior technical recruiter who spent years at Uber and Fortune 50 tech companies sat down recently to answer the unflattering questions: Is the market as bad as it looks? Is AI replacing software engineers? Are juniors actually cooked, or are companies still hiring entry-level? What niches are blowing up, and what’s quietly dying? Her answers are less catastrophic and more pointed than the timeline suggests.

📚 Table of contents

Is the market objectively bad?

Two changes are real, said the recruiter. The number of roles is down. And the quality bar companies expect from candidates is up. Combined, that means offers aren’t handed out the way they were in 2020–2021, and total compensation has tightened.

Her framing was more matter-of-fact than the social-media version: software engineering is a job where the title is “engineer” — so candidates should expect to be qualified, prep for interviews, and actually be able to code. A few years ago you could get hired with any experience signal and a pulse. That era is over. The bar didn’t get unreasonable; it got normal.

Are entry-level jobs really gone?

No — but they’re harder to find. A recent Meta posting for 200K-base entry-level roles (software engineering and product management) went viral specifically because it ran counter to the “nothing is hiring junior” narrative. The recruiter’s point: timing and visibility matter. Companies hire in waves, often without advertising broadly. If you only check FAANG career pages, you’ll miss the medtech company, the regional firm, or the unicorn that’s actively hiring this week.

Anecdotally she noted a related data point: Ford’s executives recently flagged 6,000 unfilled manufacturing roles paying ~$128K/year — not software, but a reminder that the “everything is broken” narrative is much narrower than the broader job market suggests.

What AI is actually doing to engineering hiring

Asked directly: is AI reducing demand for traditional software engineering jobs, her answer was yes — with caveats. Salesforce’s ~5,000-engineer layoff was an obvious case. Roles where AI eats the most work — marketing automation, sales-operations engineering, basic CRUD work — are hit hardest.

The flip side: AI also makes individual engineers more productive. Companies still need humans to drive projects, influence roadmaps, debug ambiguous incidents, navigate org dynamics. AI doesn’t do any of that. So the demand redistributes — less for engineers who only wrote the code, more for engineers who own outcomes.

How to talk about AI in interviews (and where it’s a red flag)

Talking about AI in interviews is now expected. Recruiters at her level are screened on how they use AI in their own workflow. If you can’t talk about how you use AI, you’re behind.

The good way to talk about it:

• I used AI to brainstorm project ideas and pick a unique angle.
• I use AI for code review and to generate small utility functions.
• I’ve evaluated which models are best for which parts of my workflow.
• I read what the AI produces critically and override it when it’s wrong.

The bad way: cheating on coding assessments with AI. Companies are actively defining AI usage policies, and the early wave of obvious cheaters has hardened detection. Use AI to prepare; do not use AI to take the test.

What the bar shift actually looks like

Five years ago: pass the coding assessment, get the offer. Now the bar layers add up:

1. Coding bar — you still have to pass it. Mock-interview-style data structures and algorithms.
2. Education signal for entry-level — degree, school, what you studied. She did not sugarcoat this: top-school CS grads still have an edge on entry-level resumes.
3. Projects with impact, not just GitHub volume — the question isn’t “did you build it,” it’s “why did you build it and what did it produce.”
4. Communication eloquence — great engineers describe their work as a chain of business impact, not as a feature list.

The most actionable lever for someone without elite credentials is #3 and #4. A small project that you can tell a coherent “here’s the user, here’s the problem, here’s what changed” story about beats a portfolio of half-finished tutorial clones.

The niches that are exploding — and the ones that are flattening

Direct from her recent hiring patterns at Uber-tier companies:

Note the framing: it’s not that front-end is dead. It’s that the absolute volume she’s seen at her companies is small compared to backend and infra. Other companies (consumer apps, e-commerce) hire front-end differently. But the trend at infra-heavy companies is real.

Why “same Python/JS/React stack” hurts you now

Every junior portfolio she reviews uses Python or JavaScript with React. That stack is fine — it’s easy to learn, ubiquitous, and useful. It’s also indistinguishable from every other bootcamp grad’s portfolio.

The stand-out move is depth in something less common:

• Java, C++, C#, Go, or Rust — languages where the supply of new juniors is genuinely smaller.
• Security — she said she has not seen a single student in her recent batches even dabble in security. Easy way to stand out.
• Infrastructure — Kubernetes, Terraform, observability, networking.
• Data engineering — pipelines, warehouses, ETL.

You don’t need a CS degree from a top school to win this round. You need a portfolio that looks unlike everyone else’s.

The recruiter’s prediction for the next 1–2 years

Two patterns will shape the near future, in her view:

• The market is cyclical. Mentors in her field describe the current downturn as “2008 all over again” — bad, then good, then bad, then good. We’re in the “bad” portion.
• It will recover, but not all the way back to 2020. Companies were “throwing cash at people who were not doing anything” in 2020–2021. That era doesn’t come back. The high bar stays; the volume of openings increases.
• AI infrastructure hiring is the floor. OpenAI alone can’t hire fast enough to fill their infrastructure positions. That pulls hiring up across the rest of the industry.
• Return-to-office complications. Three-day in-office mandates are creating real friction (people working next to copy machines because there aren’t enough desks). This affects who applies where, and may quietly reshape geographic hiring patterns.

What this means for your search strategy

Practical conclusions from the whole conversation:

• Pick a specialization that’s scarce, especially if you don’t have a top-school degree. Security and infrastructure top the list.
• Build projects you can tell a story about. Why you built it, who it serves, what changed. Not “another to-do list app for my GitHub.”
• Use AI as your idea engine and your code-review partner, not as your interview-taker.
• Apply broadly and persistently. Companies you’ve never heard of are still hiring entry-level. FAANG is one slice of one bucket.
• Prepare seriously for technical interviews. The bar is real. Two or three months of focused DSA prep + mock interviews still works.
• Communication is the multiplier. An engineer who articulates impact gets the offer over an equally capable engineer who can’t.

❌ Common mistakes the recruiter sees

Conclusion

The recruiter’s honest take is more useful than the doomscroll. The market is harder, the bar is higher, AI is real but redistributes more than it replaces. Entry-level jobs exist; they take more work to find and more polish to win. Security and infrastructure are where the demand is quietly piling up.

The advice doesn’t change with the market. Position yourself, sell yourself, prepare seriously, follow a real application strategy, and don’t give up after fifty applications. That’s the difference between candidates who land offers and candidates who post on Reddit about the death of tech.

Related reading: 7 AI engineer mistakes to avoid in 2026 — 10-step AI engineer roadmap — how to pass technical interviews in 2026

Explore More on DevShelf

• →
  Get More Software Engineering Interviews

  The practical follow-up — lane selection, LinkedIn positioning, and targeted outreach tactics that turn recruiter insight into actual callbacks.

• →
  Oracle 30K Layoffs: What It Means for Engineering Careers

  The market context behind what Uber's recruiter is seeing — why strong revenue no longer protects headcount and which roles are still growing.