“What is the state of the tech jobs market in 2025?” is the question this article tackles in the third and final part of our mini-series on that major subject. We hear from job platforms, and from tech professionals searching for their next opportunity. This article features Wellfound (a jobs platform with around 6M software engineering profiles), and data from Revealera, an alternative data platform. There are also more than 30 software engineers and engineering leaders who discussed their job hunting experience with me. Today, we cover:
Job platform data. Falling demand for remote work, more “barriers” put up by companies, slightly higher demand for backend engineers than before, and more.
Junior engineer recruitment rebounds? More scaleups and publicly traded tech companies are doubling down on hiring new grads and early-career engineers.
Picky employers. More hiring managers hold out for the “perfect candidate,” more rejections come with no feedback, and some think the candidate quality is down, overall. Referrals seem like the only way to consistently get interviews.
Software engineer archetypes in and out of demand. AI engineers, those with Big Tech experience, and infra+SRE engineers are in demand. Times are reportedly tougher for candidates who took career breaks, are self-taught, or are native mobile engineers.
State of engineering leadership hiring. It’s very tough everywhere, especially for experienced engineering managers not yet at Director level. One executive recruiter says many leadership candidates have poor AI skills and unrealistic pay expectations.
Remote market. It’s harder to land a job and compensation is lower, but the bar is higher for remote positions.
Regional observations. Outside of cities, the market is tougher: in Germany, Wayfair’s exit could drag down pay. Are Swiss companies looking to hire in cheaper EU countries?
Previous articles in this series covered:
What the data says (Part 1). Rising AI engineering demand, more Big Tech hiring, and the growing importance of location.
What hiring managers see (Part 2). A flood of inbound applications, higher demand for Product Engineers, and a challenge to find “top” candidates.
As usual, I have no affiliation with vendors mentioned in this article, and have not been paid to write about them. More in my ethics statement.
The bottom of this article could be cut off in some email clients. Read the full article uninterrupted, online.
1. Job platform data
In Part 2 of this mini-series, we covered how some hiring managers are shifting to recruit on sites like Wellfound (formerly: Angellist Talent) because LinkedIn has gotten too noisy for finding candidates efficiently.
Wellfound has 12M active candidates on its site, of which circa 50% are software engineers. More than 27,000 companies use the site for recruitment, with more than 100,000 hires made, to date. As a jobs marketplace, Wellfound does not filter. Its CEO Amit Matani has shared the trends they see, right now:
According to Wellfound’s data, the job market was most competitive in January 2025, when applications per job peaked on the site. Today, the number is down 10% from that height – meaning it’s now slightly easier to stand out as a candidate. Even so, the job market is very tough compared to 2021:
6x as many applicants per remote job
5x as many applicants per in-person job
There’s a big rise in AI engineering job ads. Overall, this is the most visible trend on the platform, and it matches trends covered in Part 1 of this mini-series. For advice on getting into AI Engineering, check out the deepdives, AI Engineering in the real world and The AI Engineering stack.
What makes an in-demand candidate?
I asked Amit about profiles of candidates which are most popular with employers, meaning types of folks whom companies reach out to the most. He revealed:
AI product engineers. Those who can build AI products end-to-end (fullstack + LLMs + evals + RAG).
Data engineers and backend engineers. Visibly rising demand for these.
AI-native engineers: Engineers who are experts in how to use AI tools to significantly boost productivity.
Pedigree means up to 50x more reachouts. Profiles with high-profile schools and workplaces get up to 20-50x more reachouts than similar profiles with not as much “pedigree”. This is not unique to software engineering, says Amit, nor to 2025.
In-person much more in demand. Most hiring is by companies headquartered in US metropolitan centers, and engineers in such areas get many more reachouts than candidates who work full-remote.
The San Francisco Bay Area has the most job postings – not even a contest. Overall, Wellfound finds the highest density of job ads in the US and India.
Remote: falling demand, but many roles are open to it
On remote job trends, Amit shares:
“35% of engineering jobs are open to remote candidates. This is down from a peak of 56% in 2022. However, remote engineering jobs get 4.5x as many applications compared to 2022.”
Visualizing this:
Hiring trends: candidates face more barriers
Below are some trends among employers from Wellfound’s data:
Hiring pipeline trends:
Inbound: noisier. This still drives some hires for all companies, but has become the noisiest channel of all. Employers face more work combing through large numbers of inbound applications.
Most companies focus on sourcing. They reach out to potential candidates directly, instead of waiting on inbound applications.
Screening trends:
More barriers for job seekers. Inbound is noisy because there’s lots of applications, but most are under-qualified. So, companies place barriers like asking additional questions upfront, scheduling async calls, assigning coding assessments – and even doing AI interviews in advance.
AI tools for ranking and filtering candidates. To deal with high volumes of inbound applications, AI tools are used to filter out applicants who are likely to be fakers, and to create an order of priority for identifying strong inbound applications. There’s still human review, and no such thing as “AI rejection”, Amit told me.
Interview trends:
More in-person interviews and lengthier takehome exercises.
Trial days or weeks, pre-offer. More companies use a trial period before making an offer. This could be a day or week where candidates work with the team. It’s pretty similar to the paid “trial week” concept at Linear.
AI is being integrated into interviews. Companies have accepted there will be AI in interviews, and are changing their processes to accommodate it.
Fake jobs, fake candidates
I asked Amit about two topics getting plenty of attention in online forums: the suspicion that some job ads are fake (meaning no hire is made and the sole purpose is to collect resumes), and that some candidates are “fake” because they use made-up CVs and hope to get hired to jobs they’re totally unqualified for.
Fake jobs: not a visible trend – but “ghosting” is. Amit shares:
“Posting fake jobs with the sole intent of capturing resumes is not widespread – at least not on our platform, or at companies I talk to. That said, companies not responding to candidates is a problem. Candidates often don’t hear back because:
Hiring managers / recruiters are indecisive or unprepared. They might have put up a role without really being ready to hire. Or they might not be sure about a candidate, but also don’t want to “close the door” by sending a rejection.
Hiring managers are overwhelmed. Some companies we work with get 1,000 applicants a week for a single role! It’s just hard to process that.
There really isn’t much of an incentive for recruiters to send a rejection email.
We try to incentivize following up: we give companies 2 weeks to accept or reject applications. We find that if the company does not respond within 2 weeks, the likelihood of them sending a followup is extremely low.”
Fake candidates are an ongoing problem. Amit:
“Fake applicants are a problem and there are three types:
Scammers or state actors who are impersonating candidates in order to deceive employers.
“Auto-appliers” who aren’t interested in a role and are a poor fit, but apply anyway.
Embellished profiles: candidates who add misleading information to their profiles; for example, stating they’re US-based when actually in India.
The biggest issue we’re seeing is scam candidates. It’s an issue for sourcing and applications. These candidates create Linkedin and Wellfound profiles, and apply to jobs. They also stuff their profiles with great companies and in-demand keywords so they show up when companies do sourcing. We make a lot of effort to keep such candidates off our platform with automated detection, reporting, and ID checks.
Auto-appliers are the next biggest thing: candidates who apply and are not good fits, overwhelm companies. A lot of them then ghost employers, even when one responds”.
More hiring trends from Wellfound
Additional observations:
Onsite hiring is dominant, but companies hiring remotely have the upper hand. Practically every startup Wellfound works with wants candidates to work onsite. But many candidates want the opportunity to work remotely, at least partially. Companies that hire remotely definitely have the upper hand in terms of talent choice.
There are AI budgets for recruitment. Many teams have “AI budgets” to try out new things with. So, there is a decent amount of experimentation in AI recruitment tools.
It’s an employer’s market. Companies know this and act accordingly: they are more picky, and put up more barriers.
Auto-appliers are growing. More candidates use auto-apply software to apply for hundreds or thousands of jobs. It’s a vicious cycle: with more inbound applications, fewer applicants hear back, and so more of them use auto-appliers.
Highest demand for software engineers with 3-15 years of experience, willing to work onsite. Amit shares that the most “promising” profile they see is for a software engineer with a few years of experience, based in locations where many companies hire. Location has suddenly become a lot more important, as reported in Part 1 of this mini-series.
Shifting demand for software engineering specializations
Revealera is a data provider of job postings for financial companies. The CTO, Henley Wing Chiu, built a machine learning model to analyze 180M job listings from the Revealera platform collected in the past 12 months. He detected trends in the subset of software engineering recruitment and has shared the findings in this article:
This chart tells us a few things:
Possible shift to fullstack engineering. The decrease in frontend engineer positions could be because of a shift to hiring fullstack engineers, instead of purely frontend engineers.
Ongoing drop in demand for native mobile engineers. This started in 2022, as covered in Is there a drop in native iOS and Android hiring at startups?
More demand for AI Engineering and Data Engineering. This is in-line with other reports, and with what we found in Part 1. Check out a deepdive on data engineering.
2. Junior engineer hiring to rebound?
One unfortunate trend since 2020 is that hiring for entry-level software engineering roles seems to have almost stopped, industry-wide:
In 2020-2022, most companies started to work remotely out of necessity due to pandemic lockdowns. Onboarding junior engineers was tough in this situation, so companies delayed hiring them.
In 2023-2024, most businesses used a hybrid setup. However, the layoff wave of 2022 increased the supply of highly skilled senior+ software engineers who used to be hard to hire. As a result, many companies hired them.
One outcome has been a very tough period for new grads and those with little professional software engineering experience. Now, there are signs that junior-level hiring may finally be returning:
OpenAI and Anthropic are hiring junior software engineers for the first time ever. I confirmed this with engineering leaders at each place while visiting their respective headquarters in San Francisco, a few weeks ago.
Shopify, GitHub, Cloudflare: uptick in internships. Shopify is hiring 1,000 interns per year, Cloudflare will be onboarding 1,111 interns in 2026 (the number is a nod to its 1.1.1.1 DNS address), and GitHub is hiring more interns than before. Only a subset of all interns convert into full-time employees, and Shopify has extended full-time offers to around 1 in 5 interns, so far. These companies have never invested as much in interns as now.
Netflix: the streaming giant only hired senior software engineers for 25 years, but has started to onboard new grads recently, as I learned by talking with Netflix CTO, Elizabeth Stone.
Below are anecdotes from hiring managers which suggest junior hiring could be picking up, based on my conversations with them:
Publicly traded tech companies and scaleups
“After two years, we’re hiring juniors again. Our company pushed AI to its limit, where people were allowed to push 10K lines of code PRs that were barely reviewed. This has caused an unmaintainable mess at the foundational layer. So my company did not hire junior engineers for the past two years. Now they are walking back from it.” – Engineering Manager, publicly traded company, San Francisco.
“We’ve started to hire juniors after 3 years. We paused junior hiring about 3 years ago, and in the summer of 2025, we still had no intern class. But the good news is that we’re targeting to hire 5-6 juniors by 2026, to bring them into a 2-year rotational path, and to get them to mid-level. We’ve started interviewing new grads more, as well.” – Director of Engineering, scaleup, New York.
“We tend to fill junior and graduate positions more easily than experienced ones. There is a larger pool of junior and graduate engineers.” – Director of Engineering, Fintech, London.
Even Amazon seems to be hiring more new grads, according to a Software Developer Manager (SDM) there in Seattle:
“For all the talk about AI replacing entry-level jobs, it feels like the trend is the opposite. We are seeing a large influx of L4s [Amazon’s entry-level role] from good schools. These new grads say most of their friends have got into Meta, Google, or similar places. There are not many left without a job”.
Pivot to hiring new grads with AI fluency
An Engineering Manager in Boston at a publicly traded company said his business is pivoting to onboarding many more junior engineers:
“We’re pivoting to hiring junior and fresh-out-of-college talent. Then, where necessary, we are hiring folks with a focus on highly scalable systems, who come from Big Tech. We’re looking to fill most open roles with very junior talent for different reasons:
Anti anti-AI bias. We’d never hire an engineer with a whiff of anti-AI bias, these days. Junior engineers are very pro-AI.
We convert the best interns to fulltime engineers. We have a very strict intern hiring approach that is not very forgiving. It has its ups and downs: we get to hire the best of the intern group, but it’s very time-consuming to give thorough feedback for everyone.
We don’t need Big Scale folks for many roles. Being hands-on with AI tools is often more important, so why hire big-scale people?
We’re investing in the bottom rung of the ladder. Junior engineers bring a lot of energy and creativity, and complement senior-heavy teams like our company has historically been.”
The same place is hiring “AI infra” engineers, the Engineering Manager revealed:
“We are investing in ‘AI infra’ with experienced engineers. We’re specifically hiring for AI engineers to embed SRE style to try and raise the benchmark of AI adoption in key areas.
There’s a lot of teeth gnashing about the death of the junior engineer, but we’re doing the opposite. I agree with the direction, and it seems to be working well, so far.”
Still tough for juniors – Wellfound
Data from Wellfound suggests the market remains challenging for junior profiles. Amit says:
“Junior folks are having a harder time. There are fewer jobs out there, and just a lot more computer science grads on the market compared to years ago. The junior characteristics that get reachouts tend to be:
Pedigree. Profiles from top schools or ‘brand name’ work experience
Practical projects. They have projects which clearly demonstrate they can build production-ready software.
AI native. Profiles that lean into being AI-native and use AI tools better. I know several companies that hire a few really senior people and a bunch of ‘AI-native juniors’ to pair with them”.
3. Picky employers
One of the most-mentioned observations is that companies are more selective with their hiring, compared to recent years. For a start, it’s hard to get responses even when candidates reply to inbound messages from recruiters. Here’s a software engineer with 13 years’ experience from Portugal, who recently worked at a crypto startup. He described the experience as “very weird”:
“90% of the time, I’m ‘ghosted’ when responding to LinkedIn recruiter messages. When I ask for more details about things like tech stack, location, if the role is remote or hybrid I don’t get any answer 90% of the time!
This experience is very strange: I never had trouble changing companies in the past, and ‘ghosting’ was not something I’ve experienced before.
To add insult to injury, when I did get to a first interview, I got rejected. This never happened in past job searches.”
Even for engineers who get lots of inbound messages, securing an offer is tough. A frontend engineer with 25 years of experience, based in California, shares:
“Hiring managers seem to be looking for the “perfect person”. I get plenty of inbound from recruiters. However, I still feel that expectations are up and salaries are down. It feels like hiring managers are in a rush to reject and not in a rush to fill positions”.
More observations by candidates:
Companies are more cautious. They seem to focus on very specific, specialized areas.
Large companies are not really hiring from startups. A software engineer in the US with 20 years of experience, and a background of working solely at startups, found very little success with reachouts. Meanwhile, his wife, a software engineer with Big Tech experience, gets far more interviews. A hiring manager said that he was the only candidate in the loop without experience at a large company, which made him feel “like showing up in crocs to a country club.” His conclusion: “I’ve felt like companies are just hoping to grab recently laid-off Big Tech workers at a discount.”
Referrals are more important than ever. Several engineers shared that their current company prioritizes referrals when fast-tracking interviews because these are more trustworthy. Meanwhile, inbound applications are stuffed with keywords, and too many unqualified candidates apply.
Long interview processes with no feedback are more common. UK-based software engineer Sam Rose shared in this thread how, after 18 hours of interviewing, he received a generic rejection and no feedback, despite going through the full interview loop.
‘Weaker’ candidates?
I asked a hiring manager who’s also looking for a job why he thinks employers are stricter about hiring. They told me they think companies are not harsher, but that candidates aren’t as strong as before:
“Candidates for our senior engineering position have been universally underwhelming. The people I’ve interviewed are either:
Well underqualified for a senior position
At a senior level in Big Tech, but an interview reveals massive skill gaps
For example, a fullstack engineer with 10 years at a big-name software company had zero experience with frontend styling systems. All engineering complexity was taken care of by a frontend platform team, or a dedicated infra team. This engineer had no in-depth knowledge of frontend systems or database internals”.
The view that candidates are weaker than before is echoed by several software engineers involved with hiring at their workplaces. Or maybe expectations have gone up!
Employers like referrals & ignore inbound applications
An experienced software engineer in the UK shares:
“Good devs with proven delivery records are approached directly, by recommendation. Lots of offers, many I’m receiving are ‘we will pay what’s necessary to get you’. I spoke with a data company last week where one of my former coworkers works, which is hiring Python devs. They received lots of CVs for inbound applications, but they are not interviewing those applicants.
In the words of the CEO: many CVs look perfect, but they can’t trust inbound candidates anymore. They would rather pay more and ‘poach’ someone they previously worked with, than pay market rate and take risks with untrustworthy candidates. So they are pinging the best engineers their team recommends and reaching out, one by one.
Recruiters working with companies are sending over multiple candidate profiles, but strongly recommend devs they worked with in the past, who they think are strong.
Trust is now the number one criterion in the UK market. For my current job, I was hired after a 15-minute chat about a project, about which I spoke barely a few sentences. This was all at a place where an engineer previously worked with me”.
Personally, I don’t think most companies completely ignore inbound applications, but there is a strong shift towards referrals. Every “fake candidate” caught in the application process will reinforce to hiring managers that they should focus on referrals and sourcing candidates directly.
Hiring dynamics during a recruitment freeze
An engineering manager at a 4,000-person company in the Netherlands shared how their company has had a hiring freeze for a year.
The company is listing “ghost jobs” due to internal politics. The company does have open positions advertised on its website, but these positions do not get approved for external hiring, even after 3 months. In reality, these are “ghost jobs”, and a bottomless pit for applicants.
This engineering manager described how the hiring freeze affects things:
Poaching from other teams is the new norm. For ICs who are high performers, there is a lot of politics, and it’s like Game of Thrones to poach from other teams because they can’t hire externally.
“Mobbing” up to push out ICs is common. EMs, Lead devs, and other management use mobbing to push out ICs they dislike. Since ICs are hesitant to look for a new job in the current job market, they look for new jobs internally. Eventually, another team hires them and they switch, meaning the externally-advertised position is filled.
Internal poaching is very important for sticking to budgets. Every team and manager is given a budget which doesn’t leave the company. If you can backfill from another team, the cost of this headcount is added to your cost center. If someone leaves your team and you cannot “poach” to backfill, then your budget will be lower next year. Poaching internally is important for those who want to be within budget by the end of the fiscal year.
Contractor rates are dropping. Teams hire contractors (freelancers) as a last resort. There’s less hiring than before, and the pay is slightly lower. Also, contract lengths used to be 12+ months – but now are no longer than 6 months due to budget uncertainty.
4. Software engineer archetypes in and out of demand
Even in the current market, certain types of engineer are incredibly in demand:
