Can HN Match Maker find your next developer?

What does HN Match Maker actually do?

HN Match Maker currently does one simple thing that Hacker News has needed for years: it tries to connect people posting in “Who Wants to Be Hired?” with companies posting in “Who Is Hiring?”

Every month, Hacker News publishes the two threads separately. Developers describe their location, technologies, remote preferences and résumé in one place, while employers describe open jobs in another. Anyone trying to connect the two has traditionally had to search through the comments manually.

HN Match Maker pulls the top-level comments from both threads and turns the free-form text into structured profiles using an LLM. It extracts skills, location, remote preferences, salary, seniority, domains and visa requirements, removes combinations that clearly cannot work, then gives the remaining candidate-job pairs a compatibility score.

The live site currently shows 105 jobs and 130 candidates. That creates as many as 13,650 raw job-candidate combinations before filtering, enough to make manual comparison tedious even though the underlying community remains relatively small.

HN Match Maker today	Count
Jobs	105
Candidates	130
Possible raw pairings	13,650

Is HN Match Maker actually proven yet?

No. HN Match Maker looks useful today, but nobody can seriously call it a proven recruiting tool yet.

The product has only just launched, so there is no public record of placements, interviews, offers, time saved per hire or successful hires that originated from one of its recommendations.

Its creator is unusually clear about that. In the Show HN discussion, he said he had no conclusions or advice to give job seekers or employers yet and wanted to observe more runs before deciding what the real matching problems were.

The launch itself has attracted real attention. When we checked, the Show HN post had reached 94 points and 40 comments. Several commenters immediately recognized the problem because they already spend time manually searching the hiring threads.

That tells us people want a better interface for HN recruiting. Whether HN Match Maker improves actual hiring outcomes is still unanswered.

Is Hacker News still a good place to find developers today?

Yes. Hacker News still produces enough serious technical hiring activity to make HN Match Maker worth building around.

The latest “Who Is Hiring?” threads continue to attract hundreds of employer comments. The previous monthly thread had 256 comments when Hacker News indexed it, while the current one had already passed 200 comments shortly after appearing.

The jobs also remain heavily technical. Recent postings include distributed systems, robotics, AI infrastructure, security, databases, Rust, machine learning, scientific computing and developer tools alongside conventional full-stack roles.

Independent index HNHIRING recently found 143 remote roles in a single monthly HN hiring thread. Its technology pages also showed 60 Python jobs, 10 Java jobs, five Kotlin jobs, four ClickHouse jobs and smaller pools for technologies such as WebAssembly and Vue.

The market has become smaller than some of HN’s busier historical hiring periods, but we still have hundreds of employers appearing every month. For a specialized technical recruiting channel, that is plenty.

Are there enough developers on HN Match Maker to make it useful?

Yes, although HN Match Maker’s candidate pool works better for targeted searches than high-volume recruiting.

The site currently contains 130 people pulled from “Who Wants to Be Hired?” posts. The underlying Hacker News thread has many more comments because discussions and replies inflate the visible count, while HN Match Maker focuses on the candidate posts themselves.

A database of roughly 130 active candidates sounds tiny beside LinkedIn. The useful difference is that these people have just publicly said they want work.

That recency changes the search. A recruiter can find thousands of developers on LinkedIn who match a technology stack and have no interest in moving. HN Match Maker starts with a pool that has already declared hiring intent.

For a company trying to recruit 50 engineers, the pool is obviously too narrow. A startup looking for one senior systems engineer, a founding developer or someone with an unusual mix of infrastructure and domain knowledge has a much more plausible use case.

What kind of developers can you actually find through HN Match Maker?

HN Match Maker is strongest today for software engineering roles where technical depth matters more than sheer candidate volume.

A quick look at the live jobs shows companies searching for Rust, C++, Go, TypeScript, React, PyTorch, AWS, PostgreSQL, networking, machine learning, distributed systems and infrastructure experience.

The current listings also include much narrower work. Fastly is hiring across edge computing, networking and security. Monumental wants software engineers working on autonomous construction robotics. Steg.AI is looking for an engineer capable of productionizing deep-learning watermarking models. Other recent HN postings involve quantitative trading systems, high-performance storage, OpenZFS, embedded engineering and scientific computing.

That mix explains why Hacker News has remained useful despite its small size. Many people reading HN already work close to software infrastructure, developer tools, startups or technically demanding products.

HN Match Maker becomes especially interesting when a company needs someone harder to describe with a generic title such as “software engineer.”

Does Hacker News really lead to developer hires?

Yes. Hacker News has produced enough real developer hires over many years that the underlying talent source is no longer the speculative part of HN Match Maker.

In a 2022 Ask HN discussion, one hiring manager said he had hired two people from the monthly threads and both were doing very well. Another said he had hired two part-time employees and considered HN the highest-quality traditional job-listing channel he had used.

A different hiring manager described searching three months of “Who Wants to Be Hired?” posts, narrowing the pool to seven frontend developers and completing the hire within roughly one or two weeks.

Developers told similar stories. One person said 25 applications generated six final-round processes and two offers. Another said he had found two jobs through HN. Others reported staying four or five years at companies they originally discovered through the hiring threads.

More recent discussions show the channel still works. Earlier this year, one developer said he had found two jobs through HN and turned down a third offer, estimating roughly a 50% success rate on his highly targeted applications. Another recent user said HN had led to three hires and eight years of employment in total.

These are self-reported outcomes rather than a controlled hiring dataset, so we cannot calculate a meaningful overall conversion rate. We can be much more confident about the basic point: employers and developers have repeatedly found each other through these threads.

How does HN Match Maker decide which developer fits a job?

HN Match Maker currently gives most of its weight to technical skill overlap, then checks domain experience, seniority and compensation.

The process starts by extracting structured fields from both Hacker News posts. Before scoring, HN Match Maker removes combinations where the basic employment conditions conflict, such as a developer requiring sponsorship when the company explicitly cannot provide it.

The remaining combinations receive a score from zero to 100. Skills make up 40% of the result, domain fit 20%, seniority 20%, salary overlap 10% and a common base score the final 10%.

That weighting makes sense for an initial shortlist. A company searching for a senior Rust infrastructure engineer should care far more about relevant engineering experience than whether two people happened to phrase their posts similarly.

The harder question is whether HN Match Maker can recognize that relevant experience reliably from very short public posts. Its first run shows that this is where most of the work remains.

HN Match Maker score	Weight
Skill overlap	40%
Domain fit	20%
Seniority	20%
Salary overlap	10%
Base score	10%

Why are HN Match Maker's scores so low right now?

HN Match Maker’s scores are low because employers and developers write their Hacker News posts in almost opposite ways.

The creator noticed this immediately while reviewing the current matches. Employer posts tend to be very specific. A company may describe a particular domain, exact technologies, seniority, location, compensation and several responsibilities.

Candidates usually stay broad. They might write “Python, Go, PostgreSQL, AWS” and mention ten years of engineering experience without explaining every industry they have worked in or every production system they have built.

That makes sense for someone advertising themselves once to hundreds of employers. A highly specific candidate post would reduce the number of companies likely to make contact.

The mismatch has a measurable consequence. According to the creator, every pairing he inspected in this first run remained below his 50% heuristic threshold.

That tells us more about the available text than about the quality of the developers. A candidate may have exactly the right experience while leaving the relevant domain or project out of a deliberately short public post.

Should you trust an HN Match Maker score?

No. HN Match Maker scores are useful for sorting candidates today, but treating them as hiring grades would be a mistake.

The site itself says scores measure plausible fit from the public text and should serve as a starting point for conversation. That is an important limitation because the underlying profiles are incomplete.

Consider a developer who writes “Go, Kafka, Kubernetes, PostgreSQL” after years of building high-throughput infrastructure. An employer might describe a role using phrases such as “large-scale event processing,” “distributed data systems” and “streaming architecture.” A human engineer can see the overlap quickly. A tag-based system can miss part of it.

One commenter in the launch discussion noticed this exact weakness after finding very specific extracted tags such as “large-scale data systems.” He suggested mapping posts into a controlled vocabulary. The creator said embeddings were another possible way to improve the matching.

That is probably the biggest technical opportunity for HN Match Maker. Better semantic matching could move the product beyond counting shared words and closer to recognizing shared experience.

Until then, we would use a 45% match to decide which profile to open first. We would never interpret 45% as a probability that the person will succeed in the job.

Can HN Match Maker save time on remote, salary and visa mismatches?

Yes. HN Match Maker may already save more time by removing impossible matches than by perfectly ranking the remaining developers.

Remote work on Hacker News is full of hidden restrictions. One job can say remote and mean United States only. Another accepts North America. A European company may require EU working hours. Some companies support relocation or sponsorship while others explicitly do neither.

Recent HNHIRING listings show the range clearly. Portless advertised remote work limited to North America. One enterprise software company was hiring remotely across a defined list of Latin American countries. Sequent Tech advertised a global remote role. Other companies mix remote and onsite hiring depending on the team.

HN Match Maker extracts location, remote preference, compensation and visa needs, then uses some of these fields to remove impossible combinations before scoring.

Those are valuable filters because they deal with facts rather than fuzzy judgments about engineering ability. If a candidate needs sponsorship and the company cannot sponsor, a brilliant skills match still goes nowhere.

A recent HN discussion accidentally showed why this matters: one developer said an employer contacted him, asked what country he lived in and then rejected him because of that country, even though his location had already been written in the original “Who Wants to Be Hired?” post.

Better filtering can prevent exactly that kind of wasted conversation.

Can HN Match Maker find developers you would miss with Ctrl-F?

Yes. Finding non-obvious matches is the part of HN Match Maker that could eventually make it much better than manually searching the monthly thread.

Ctrl-F works surprisingly well for obvious requirements. A developer looking for Rust can search “Rust.” A Toronto candidate can search “Toronto.” One developer who later got hired through HN described essentially doing that.

The method gets weaker when the same experience has several names.

A company hiring for distributed systems may want someone who describes Kafka, high-throughput backends, Go and Kubernetes without ever writing “distributed systems.” A robotics company may care about embedded Linux, computer vision and hardware integration even if “robotics” never appears in the candidate's technology list.

HN Match Maker can theoretically connect those descriptions because it extracts skills and broader domains before comparing them.

Its current tag system still misses some semantic overlap, as the launch discussion revealed. If that improves, this becomes the product's clearest advantage: surfacing the candidate a recruiter would never have thought to search for.

Can candidates game HN Match Maker with keywords?

Yes. HN Match Maker’s current scoring system gives candidates a clear incentive to stuff more technologies and domains into their posts.

Skills account for 40% of the score, so broader technology lists can create more apparent overlap with job descriptions. The creator has already acknowledged this problem and said candidates could game the ranking by dumping keywords into their posts.

Hacker News historically had less incentive for this behavior because “Who Wants to Be Hired?” posts were written mainly for human readers. A good engineer could simply write five relevant technologies and let an interested employer open the résumé.

Visible algorithmic rankings change that incentive. Anyone who understands the scoring system can start writing for the matcher.

A mature version of HN Match Maker would need to distinguish real evidence from vocabulary. Five years building production Rust systems should count for much more than adding “Rust” to a long list of 30 tools.

Today, that distinction still requires a human reading the original post and résumé.

Can HN Match Maker tell whether a developer is actually good?

No. HN Match Maker currently knows enough to estimate fit, while actual developer quality still has to be evaluated elsewhere.

The “Who Wants to Be Hired?” format typically contains location, remote preference, willingness to relocate, technologies, a résumé link and contact information. Some candidates add years of experience or a short explanation of what they have built.

That is enough to answer useful early questions. Does this developer know the relevant stack? Is the seniority roughly right? Can the person work in the required location? Is there obvious domain overlap?

It says much less about code quality, communication, judgment, reliability, leadership or how deep any particular skill goes.

The current HN Match Maker profiles make the gap easy to see. One candidate profile, for example, contains 15 years of experience alongside Python, C++, ClickHouse, Kafka, PostgreSQL, AWS, PyTorch, reinforcement learning, trading systems, distributed systems and FPGA integration. That is enough to make the person interesting for several specialized jobs. We still have no idea whether that person should be hired without examining the actual work and talking to them.

HN Match Maker can shrink the search space. Technical interviews, references, project review and normal hiring judgment remain essential.

Is the privacy criticism around HN Match Maker a real problem?

Yes. HN Match Maker has already hit a genuine privacy tension because aggregating public candidate information changes how easily that information can spread.

The launch discussion became surprisingly heated around this point. One commenter who had considered posting in “Who Wants to Be Hired?” described the public matching interface as invasive. Another argued that people deliberately place this information on a public forum so potential employers can find them.

The strongest concern is practical rather than philosophical. Several HN users described getting spam after posting contact details publicly. One person in the current discussion said a hidden email address had received thousands of DocuSign messages. Another said an old job post kept generating approaches months after the position had disappeared because other sites had copied it.

Hacker News already tells readers to contact people in the candidate thread only about work opportunities. HN Match Maker can make relevant outreach easier, although the same aggregation can also make indiscriminate outreach easier.

For employers, the rule is simple: use the matcher to send fewer, more relevant messages. Turning 130 public candidate profiles into another bulk-email list would quickly destroy the part of HN recruiting that makes it useful.

Is HN Match Maker actually better than the other Hacker News job-search tools?

HN Match Maker currently adds something useful that most HN job indexes do not: it compares the employer and candidate threads directly.

Hacker News already points job seekers toward several search interfaces. HNHIRING lets people browse historical and current roles by technology and location. Other tools such as nthesis.ai and wantstobehired.com make one side of the monthly hiring market easier to search.

Those tools work well when the user knows what to type.

HN Match Maker tries to remove that step. A developer profile can automatically be compared with every employer post, while an employer can browse candidates ranked against a specific job.

That changes the workflow from “search the database” to “show me who seems relevant.”

Whether HN Match Maker's rankings are already better than a smart human running a few searches is doubtful. The low first-run scores and messy extracted tags show why.

The more interesting comparison will come after the semantic matching improves. If HN Match Maker repeatedly surfaces candidates that employers failed to find with ordinary keyword search, it will have a real advantage over existing HN interfaces.

Should a startup use HN Match Maker to hire a developer now?

Yes. A startup hiring a technical role should already try HN Match Maker, especially when one strong candidate matters more than a huge applicant funnel.

The cost of checking it is tiny. The candidates are actively advertising their availability, the pool is heavily technical, and the matcher can quickly expose people who share relevant skills, domains or working constraints.

We would pay particular attention to founding engineers, senior backend developers, infrastructure specialists, developer-tool engineers, ML engineers and other roles where Hacker News tends to have unusually relevant people.

We would also deliberately inspect candidates below the very top score. The current scoring model still depends heavily on how much detail someone chose to write publicly, and the creator's first results show that the numbers remain compressed.

Posting directly in “Who Is Hiring?” still makes sense too. Historical HN discussions repeatedly show that employers can get strong inbound candidates from the thread itself.

HN Match Maker adds another route into the same talent pool. For a startup that only needs one excellent developer, another good route is worth having.

Can HN Match Maker really find your next developer?

Yes, HN Match Maker can realistically surface your next developer, although today the tool is much better at finding people worth talking to than predicting who you should hire.

The strongest part of the case comes from Hacker News itself. Developers and employers have been finding each other through the monthly hiring threads for years, and recent users are still reporting successful hires and strong response rates.

HN Match Maker removes one of the worst parts of that process: manually comparing two large piles of unstructured posts. It already extracts the useful fields, removes some impossible combinations and gives employers a manageable place to start looking.

The weak point is the ranking. Candidate posts are often too general, employer posts are unusually specific, semantic tags remain messy and the current scores are weak enough that they should never be mistaken for hiring probabilities. As seen above, the creator openly acknowledges these limitations.

That still leaves a useful product. If HN Match Maker shows a founder five developers they would otherwise have missed, and one turns into a great hire, the algorithm does not need to solve recruiting.

For now, our answer is yes: HN Match Maker is worth using to find your next developer. Just use it as a discovery tool, because the evidence that its ranking can reliably identify the best candidate has not arrived yet.