We Analyzed 35,000 AI Startup Job Listings. Only 1.3% Are Entry-Level.
By Fast AI Startup Jobs
The AI job market is booming. Funding is pouring in. Every week another startup announces a new model, a new product, a new funding round. And if you're early in your career, you're probably wondering: where do I fit in?
We ran the numbers. The answer is uncomfortable.
TL;DR / Key Takeaways
- Only 458 out of 35,450 AI startup jobs (1.3%) are explicitly tagged as intern or new-grad level. The rest are listed as experienced — though "experienced" is a crude label, and many of those roles may only require 1–2 years of relevant work. The real accessible number is likely higher, but the data can't tell us by how much.
- Software engineering is the single largest entry-level category — 197 of the 458 roles, or 43% of all entry-level openings.
- Early-stage companies (Series A, Seed) are your best shot. They're 3–5x more likely to hire entry-level than late-stage or growth-equity-backed companies.
- 76.9% of "entry-level" roles are full-time positions, not internships — meaning companies do want to hire permanent junior staff, just very selectively.
The Full Picture: What 35,000 AI Startup Jobs Actually Look Like
Our dataset covers 35,450 job listings across 1,542 AI startups, gathered from the Fast AI Startup Jobs dataset of active AI companies. These aren't legacy enterprise jobs or Big Tech roles. These are the startups building at the frontier — the companies that job seekers most want to break into.
Here's the headline finding:
| Category | Count | Share |
|---|---|---|
| Experienced (any level) | 34,244 | 96.6% |
| Entry-level / Intern / New-Grad | 458 | 1.3% |
| Unlabeled / Unknown | 748 | 2.1% |
One point three percent. That number deserves scrutiny — and context.
It does not mean 98.7% of AI startup jobs are inaccessible to early-career candidates. "Experienced" is a crude default label in most ATS systems, applied to everything that isn't explicitly flagged as an internship or new-grad program. Many roles tagged "experienced" may only require 1–2 years of relevant work. The real number of accessible jobs is likely higher — but the data can't tell us by how much.
What the 1.3% does reveal is this: AI startups are not building structured junior hiring pipelines. The dominant model is "hire seniors, move fast, ship fast." Formal entry-level programs — the kind with cohort onboarding, mentorship tracks, and rotational assignments — are rare. When juniors get hired, it tends to happen opportunistically, not systematically.
This doesn't mean it's impossible. It means you need to be strategic in a way that most job seekers aren't.
Where the 458 Entry-Level Jobs Actually Are
Not all entry-level opportunities are created equal. The breakdown by department reveals where companies are actually willing to take chances on junior talent.
| Department | Entry-Level Jobs | Share of 458 |
|---|---|---|
| Software Engineering | 197 | 43.0% |
| General / Unclassified | 91 | 19.9% |
| Research | 31 | 6.8% |
| Marketing & Growth | 22 | 4.8% |
| Analytics | 20 | 4.4% |
| Sales & Partnerships | 15 | 3.3% |
| Infrastructure & Platform | 10 | 2.2% |
| Recruiting & Talent | 10 | 2.2% |
| ML / AI Engineering | 9 | 2.0% |
| Design | 8 | 1.7% |
| Data Science | 8 | 1.7% |
Software Engineering Dominates — But There's a Catch
Nearly half of all entry-level openings are in software engineering. This makes sense: SWE roles have the most mature hiring pipelines for junior candidates, structured onboarding playbooks borrowed from FAANG, and well-defined career ladders.
But note what's nearly absent from this list: ML/AI Engineering has only 9 entry-level slots across 35,000 jobs. Data Science has 8. If your plan was to land a junior ML engineer role at an AI startup right out of school, the data says that path is extremely narrow — not impossible, but narrow.
The "General / Unclassified" Category Deserves Attention
91 roles (19.9%) are unclassified. These are jobs where the company didn't bother with department tags — often early-stage startups posting their first few hires. This is actually a signal worth tracking. Unstructured job listings at young companies frequently represent the most flexible roles, where the actual responsibilities are still being defined and a strong candidate can shape the position.
Non-Technical Roles Are Underrated
Marketing & Growth (22), Analytics (20), Sales & Partnerships (15), and Recruiting (10) together account for about 15% of all entry-level openings. These roles are often overlooked by candidates fixated on engineering or research. For context, our analysis of the full dataset shows that 46.6% of all AI startup jobs — not just entry-level — sit outside engineering departments. The non-technical side of AI is far larger than most people assume, and entry-level openings exist there too.
Funding Stage Matters: Early-Stage Companies Are More Open
This is the most actionable finding in the data.
We looked at the percentage of each company's job listings that were tagged entry-level, broken out by funding stage:
| Funding Stage | Entry-Level % of Total Jobs |
|---|---|
| Series A | 2.1% |
| Seed | 2.0% |
| Series B | 1.8% |
| Series F | 1.5% |
| Series E | 1.2% |
| Series G+ | 0.7% |
| Series C | 0.6% |
| Series D | 0.5% |
| Growth (PE / Late) | 0.4% |
| Pre-Seed | 0.0% |
| Bootstrapped | 0.0% |
The pattern is clear: Series A and Seed companies hire juniors at roughly 3–5x the rate of late-stage companies.
The logic isn't hard to follow. A Series A company with 15–40 employees has figured out its core product direction but still needs to scale its team fast. Juniors are cheaper, easier to mold to the company's culture, and represent a long-term talent investment. A Series D company with 300+ employees, on the other hand, is optimizing for execution speed — they want people who can contribute from week one, no runway required.
Two findings at the extremes are worth noting. Pre-seed and bootstrapped companies show 0.0% entry-level hiring. Pre-seed companies are typically 2–5 founders who need every hire to carry an unrealistic amount of weight. Bootstrapped companies, by definition, have limited capital and can't absorb the cost of junior onboarding. Don't target these for your first AI job.
The practical implication: filter your job search to Series A and Seed companies. They're harder to find than the Anthropics and OpenAIs of the world, but the math is firmly in your favor by comparison.
The "Experienced" Label: How Strict Is It Really?
Here's an important caveat the raw data can't fully answer: what does "experienced" actually mean in a job posting?
In many cases, "experienced" is applied to any non-intern role as a default — including positions that only ask for 1–2 years of relevant work in the requirements. Companies frequently use "experienced" as a label without meaning "we want 5 years." Conversely, some jobs tagged "entry-level" come with skills requirements that would challenge a mid-career professional.
The label is a filter, not a law. Job descriptions that list 2 or fewer years of experience, or use language like "strong fundamentals," "passion for X," or "ability to learn fast" are often genuinely accessible to early-career candidates — even when filed under "experienced" in a database.
This means the real number of AI startup jobs accessible to early-career candidates is likely higher than 1.3%. But by how much? Impossible to say without reading every job description. What the data does confirm is that startups are not building junior pipelines at scale. When entry is possible, it tends to happen through direct outreach, referrals, or unconventional positioning — not the standard application queue.
What the Data Suggests You Should Do Differently
Three findings from the data translate directly into job search strategy.
1. Filter for Series A and Seed Companies
Your base rate of finding an entry-level opening is roughly 2% at Seed/Series A versus 0.4–0.7% at late-stage companies — a 3–5x difference. Target companies funded within the last 6–18 months. They're in hiring mode, and their entry-level odds are the highest in the dataset.
2. Lead with Software Engineering
197 of 458 entry-level openings (43%) are in SWE. If you have any engineering background — even a bootcamp or strong portfolio — this is your widest funnel. Once inside, pivoting toward ML, data science, or product is far easier than breaking in cold.
3. Don't Overlook the 91 "Unclassified" Listings
These roles are disproportionately at earlier-stage companies posting their first hires in a given function. They're less competitive (fewer eyes), more flexible in requirements, and often closer to the founding team. They are the hidden inventory of the entry-level AI job market.
For non-technical entry points — marketing, analytics, sales, recruiting — see our companion analysis: 47% of AI Startup Jobs Aren't Engineering Roles.
Methodology
The data in this article comes from the Fast AI Startup Jobs dataset of 35,450 AI startup job listings across 1,542 active companies. Jobs were classified as "entry-level" or "intern/new-grad" based on job title tags, seniority labels, and listing metadata. "Experienced" was assigned to all other roles. Funding stage data was sourced from company profiles in the same dataset. The snapshot reflects job postings collected as of early 2026.
The department breakdown covers all 458 entry-level jobs; percentages are rounded to one decimal place and may not sum to exactly 100%.
Limitations to keep in mind: The dataset is sourced from major ATS platforms (Ashby, Greenhouse, Lever, and others). Companies that post jobs only on their own websites, or use ATS platforms we don't track, are not represented. This means the data skews toward companies that have adopted structured hiring tools — generally companies with 10+ employees. Very early-stage teams hiring via Twitter DMs or personal networks are invisible here.
Additionally, job-level tagging is inconsistent across companies, and the "experienced" label is applied inconsistently by employers. The 1.3% figure represents explicitly tagged entry-level openings — the true accessible population may be somewhat higher. Analysis of job description text requirements was outside the scope of this dataset query.
Data sourced from the Fast AI Startup Jobs dataset. Analysis reflects 35,450 listings across 1,542 companies.
Related: 47% of AI Startup Jobs Aren't Engineering Roles. Here's the Data. — Our analysis of the same dataset shows that nearly half of all AI startup roles sit outside engineering departments.