How Early-Stage Startup Recruiters Are Solving Their Hardest Hiring Problems Right Now
By
Samantha Cox
•
We sat down with technical recruiters from five VC-backed startups (Blacksmith, Estuary, Manifest Cyber, Revive, and Talkiatry) at our latest AI & OJ event and asked them to stop networking and start comparing notes. What followed was one of the most tactical recruiting conversations we have hosted. Here is what came out of it.
Candidate Fraud Is Costing Recruiters Hours Every Week
Fraud was the first topic on the table and the one that got the most immediate agreement. Every recruiter in the room had dealt with it, and several had seen fraudulent candidates make it all the way to the offer stage before being caught.
The most common fraud patterns recruiters are seeing at early-stage startups:
Deep fakes and fabricated identities where the person on the video call does not match the LinkedIn photo
Stolen LinkedIn profiles where someone applies using another real person's credentials and work history
Keyword-stuffed resumes that look technically perfect on paper but fall apart under any questioning
AI-generated take-home submissions, where candidates produce polished deliverables but cannot walk through their own work on a follow-up call
Remote software engineering roles see the highest fraud rates. One recruiter was manually verifying phone carriers before every screen call to confirm candidate identities before their company switched to Ashby, which automates much of that filtering.
Fraud is not just a top-of-funnel problem anymore. Take-home assignments that used to take candidates a week are now coming back in two days, and the quality looks good on the surface. The real verification happens in the live follow-up conversation, which means recruiters need to design their processes with the assumption that written deliverables alone cannot be trusted.
The Hybrid Role Problem Is Getting Worse
Several recruiters are stuck on roles that do not fit neatly into traditional job categories. This is not a new challenge, but it is intensifying as engineering disciplines merge and startups ask individuals to cover more ground.
Role | The Challenge | Why It Is Hard |
|---|---|---|
Solutions Engineer (Series A, data platform) | 70% technical support, 30% strategic product thinking, 3 to 5 years experience | Open since December. Candidates tend to be strong on the technical side or the strategic side, rarely both. A previous hire interviewed well but turned out to be a completely different person on the job. |
Director of Sales (e-waste recycling tech) | Titled as director but scoped at manager level, Brooklyn Navy Yard location requirement | Director title attracts overqualified candidates. The right candidates often do not apply to director-level postings. Physical location narrows the pool further. |
Director of Engineering (e-waste recycling tech) | Needs VP-level thinking but hands-on execution at a small company | At smaller companies, leadership roles still require daily individual contribution. Most director candidates expect to manage, not build. |
Design roles (telehealth) | Increasing overlap between design and front-end engineering | The line between designer and front-end developer is blurring. Finding people who sit at that intersection is harder than hiring for either discipline separately. |
The broader trend the group identified is that traditional role boundaries are dissolving. Backend engineering, data engineering, and systems work are merging into single positions. Design and front-end development are doing the same. The job descriptions that worked two years ago do not always reflect what companies actually need from a hire today, and that mismatch is showing up as longer time-to-fill and more interview cycles that go nowhere.
A Sourcing Strategy That Gets a 50%+ Response Rate for Staff and Principal Engineers
The most detailed sourcing playbook came from a recruiter hiring staff, and principal-level engineers at a Series A bare metal cloud company focused on AI/ML infrastructure. The approach uses Claude as a research layer and the founding team as the engagement mechanism.
How it works:
Build a shortlist of target candidates based on the technical profile
Load the shortlist into Claude with a prebuilt research prompt
Claude identifies where each candidate is most active online (research papers, blog posts, conference talks, social media, open source contributions)
Claude ranks each candidate's activity by platform so the recruiter knows where engagement will land best
The recruiter shares the top candidates and their content with the founding team
Founders leave genuine, thoughtful comments on the candidates' published work (blog posts, papers, open source repos)
Only after the candidate has seen the company name in an organic context does the recruiter send outreach
The response rate with this approach is above 50%, which is remarkable for staff and principal-level engineers who typically ignore cold outreach entirely.
Why this works is worth understanding. Principal engineers do not move companies. They move for problems they want to solve. Cold InMails and generic job descriptions do not signal that you understand what a senior engineer cares about. But when a founder comments on their blog post or research paper with a substantive take, that tells the candidate something about the company's technical depth before anyone has even pitched the role.
This approach requires more upfront effort per candidate, but the yield is dramatically higher. And for principal-level hires where a single person can change the trajectory of a company, the math makes sense.
Culture-First vs. Technical-First Screening
The group split on whether to frontload culture fit or technical ability in the interview process. The answer mostly came down to two variables: team size and whether the role is collaborative or individual-contributor focused.
When culture-first screening works best:
Small teams (under 10 to 15 people) where everyone works closely together daily
Roles that involve heavy cross-functional collaboration
Companies where the hiring manager has strong opinions on team dynamics
One recruiter immediately rejects candidates who have not researched the company before the recruiter call
Another treats it as a hard no if an engineer says they "leave work at work" and have no interest in engineering outside their job
When technical-first screening works best:
Larger teams or organizations where IC depth matters more than team chemistry
Outbound sourcing for senior engineers, where pushing company enthusiasm too early can feel like a sales pitch and turn candidates off
Roles where the technical bar is the primary filter and culture alignment is evaluated later
The interest in the company often develops over time for senior candidates as they learn about the problems, not because a recruiter tested for it in the first call
The function also matters. Sales candidates are expected to research the company before a screen. Engineering candidates generally get more leeway because the technical evaluation carries more weight at that stage.
The takeaway is not that one approach is better. It is that the decision should be intentional and based on the role, not just defaulted to because "that is how we have always done it."
How Knockout Rates Reveal Whether Your Funnel Is Working
Knockout rates at the recruiter screen varied significantly across the group, and the differences were revealing.
Recruiter Profile | Knockout Rate at Recruiter Screen | What It Signals |
|---|---|---|
Hiring staff/principal engineers at Series A, very specific ICP | 70 to 80% | High bar, tight profile. Most candidates do not fit the profile of someone who will take a company from seed to Series D. This recruiter sees the high knockout as the system working. |
Cross-functional roles at telehealth company, mix of corporate and tech | 40% | Moderate filter with a very high pass rate at the hiring manager stage. The recruiter screen is sending the right people through. |
Solutions engineer at Series A data platform, focus on vibe and interest | Low (estimated under 20%) | First round functions more as a vibe check than a filter. The technical screen becomes the real bottleneck. |
The insight the group converged on: if your technical screen is the biggest choke point in your pipeline, the fix is usually upstream. Tightening the recruiter screen so that fewer unqualified candidates reach the technical round is almost always more efficient than redesigning the technical interview itself.
One recruiter framed it well. A high knockout rate at the top of the funnel is not a sign of a broken process. It is a sign that the recruiter knows the profile well enough to filter efficiently, which protects the time of hiring managers and engineers who are doing the downstream interviews.
A Two-Hour Technical Interview That Tests What Actually Matters
The most detailed interview process came from the recruiter at Blacksmith, a Series A bare metal cloud company. Their process is designed to test three things: technical depth, product thinking, and communication.
Stage 1: 30-minute recruiter screen
Two questions carry almost all the weight:
What do you know about Blacksmith? If the answer is nothing, it is an immediate rejection. For a company hiring people who will shape the product direction, zero research signals zero interest.
What is the toughest thing you have built? This reveals depth, ambition, and whether the candidate has operated at the level the role requires.
This recruiter talks to about nine candidates a day and maintains a 70 to 80% knockout rate at this stage.
Stage 2: Two-hour collaborative technical interview
Candidates build a background coding agent, which is a feature Blacksmith has already built internally. They receive a starting point and need to complete the build within the session. The interviewer evaluates:
Product obsession: Does the candidate have strong opinions on what good UX looks like? Do they make intentional design choices or just get something functional?
Building approach: The rubric has two tracks. Linear builders work step by step. Non-linear builders jump to the hardest problems first and fill in gaps later. Both approaches can produce excellent results, so the rubric evaluates each on its own terms rather than penalizing one style.
Communication: After building, the candidate demos what they created. This tests whether they can explain technical work to non-technical stakeholders, which matters for a small team where engineers interact directly with business functions.
Systems design: One standard question accounts for about 25% of the rubric and provides a consistent comparison point across all candidates.
The collaborative format is intentional. Rather than watching a candidate solve a problem on a whiteboard, the interviewer works alongside them, which reveals how someone actually operates as a teammate.
Time Management for Solo Recruiters Handling Both Sourcing and Screening
Several recruiters in the room are the only recruiter at their company or part of a two-person team. The question of how to protect time for sourcing when your calendar is full of calls came up, and two different approaches stood out.
Approach 1: Full-day batching
One recruiter dedicates entire days to a single function. Recruiting days are for back-to-back candidate calls (up to nine per day). Sourcing days are for deep research, outreach sequences, and strategy with zero calls on the calendar. The key is eliminating context switching, which this recruiter found was the biggest productivity killer when trying to mix sourcing and calls in the same day.
Approach 2: Protected time blocks
Another recruiter blocks three one-hour windows throughout the week specifically for sourcing and admin. Everything outside those blocks is calls and active candidate management. The windows are non-negotiable and do not get bumped for scheduling convenience.
Both approaches share the same underlying principle, which is that sourcing will not happen unless it has dedicated, protected time on the calendar. If it is treated as something to squeeze in between calls, it does not get the focus it needs and pipeline quality drops.
How Top Recruiters Are Closing Senior Engineers
The conversation closed on candidate experience, and the tactics that came up went well beyond the standard offer letter and benefits PDF.
Personalized offer pages built in Webflow
One recruiter builds password-protected, personalized web pages for final-stage senior candidates. Each page includes:
The specific technical problems the company is solving and which ones align with the candidate's interests
A detailed equity breakdown with explanations of early exercise options
Benefits tailored to the individual offer, not a generic company-wide summary
These pages serve two purposes. They give the candidate real substance to evaluate instead of a generic offer letter, and they give them something tangible to share with a partner or financial advisor when making the decision. For senior engineers weighing multiple offers, this level of transparency is a differentiator.
Other closing tactics the group discussed:
Day-in-the-life content that shows what working at the company actually looks like on a Tuesday afternoon, not just the highlight reel
Dynamic landing pages for specific roles that highlight the internal problems the team is solving
Employee value proposition materials that go deeper than standard job descriptions
Long-term relationship building with principal-level candidates, sometimes cultivating a hire over a full year before the candidate is ready to move
The common thread is that the best recruiting at the senior level feels less like selling and more like giving someone the information they need to make a clear-eyed decision.
The Recruiting Tools That Came Up Most
Claude was the most discussed tool by far. Usage ranged from simple individual workflows (rewriting outreach messages, summarizing candidate profiles) to fully integrated systems where Claude is connected to a company's candidate database, Google Drive, and Slack, allowing recruiters to query any information about candidates or companies in natural language.
Tool | How Recruiters Are Using It |
|---|---|
Claude | Candidate research and activity ranking, messaging and outreach sequencing, workflow automation, dynamic landing page creation, full data integration with candidate databases and internal tools (Google Drive, Slack) |
Ashby | ATS with built-in fraud detection and filtering, replaced manual phone carrier verification for one team |
Juicebox | Sourcing and candidate discovery |
Gem | Outreach sequencing and candidate engagement tracking |
Clay | Outreach automation and data enrichment |
FAQ
What is AI & OJ?
Who attends AI & OJ events?
How do I get invited to AI & OJ?
What recruiting tools are startups using right now?
How are recruiters using AI in hiring?
