AI in recruiting: What Q2 Tells Founders, VCs and the Future of Talent

This post summarizes a wide-ranging conversation on This Week in Startups with three investors — Astasia Myers (Felicis), Matt Turck (Firstmark), and D.A. Wallach (Time BioVentures) — after we closed Q2 and started Q3. The episode covered valuations, deal velocity, biotech’s “great depression,” the explosion of generative AI, and how founders should approach hiring, runway and go-to-market. One recurring theme that matters to every startup and hiring leader is AI in recruiting, and the ways this technology is reshaping how teams are built, how talent is found, and how companies scale.

Why this quarter matters: macro snapshot and what we discussed

After Q2 closed, the world felt like it was turning a page. Public markets were giving mixed signals, M&A activity picked up, and private-stage behavior shifted dramatically. Three beats stood out during our talk: first, early-stage rounds were fewer but larger; second, AI-fueled companies were raising at record prices; and third, biotech remained deeply challenged with its public exit window largely closed. All of this creates a new operating environment for founders and investors. For founders thinking about hiring, a major takeaway is how quickly AI in recruiting is changing expectations — from the size of founding teams to how you scale your first 20 people.

Early-stage pricing pressure: high valuations, selective investing

Astasia and Matt both described the same paradox: the quality of founders is improving and the total addressable opportunities have widened, but valuations at Series A are higher than either of them has seen in their careers. Astasia emphasized that the "AI explosion democratizes access to tools," enabling teams to move faster and achieve ARR earlier in their lifecycle. Matt was blunt: interesting AI companies routinely carry nine-figure pre-money valuations today at Series A — a reality that makes entry pricing tough.

For founders, this sets up a discipline challenge: when should you accept capital, and how much should you raise? The answer is not one-size-fits-all, but smart founders are focusing on cash efficiency. Being lean and deliberate with capital — modeling burn, runway, and the use of funds — is back in vogue. That shift also interacts with hiring: many founders are taking a different approach to early headcount to preserve runway and invest in distribution. This is yet another area where AI in recruiting enters the conversation, because tools are enabling smaller headcounts to accomplish more.

Biotech's "Great Depression": why it matters

D.A. framed biotech as being in a multi-year slump brought on by the absence of public capital. Historically, biotech has leaned on public markets as a capital formation mechanism — companies often go public to access the funding required for clinical trials. When the public door is closed, the private markets feel the squeeze.

What happened? Generalist public investors rotated out of biotech, leaving only smaller specialist funds. Without large generalists driving IPO and follow-on demand, many companies are stuck waiting. Even biotech companies waving an AI banner haven't fully broken the malaise, since many specialist investors remain cautious. For founders building in biotech, this means a longer, more capital-constrained road. For recruiting, it means hiring strategies that favor outsourcing and leveraging contractors or CROs (contract research organizations) rather than large FTE counts early on.

AI isn’t just another software wave — it threatens labor budgets

One of the most provocative parts of the conversation was Astasia’s framing of AI as a platform shift that targets labor budgets rather than simply software budgets. Where prior waves (on-premise → cloud, desktop → mobile) often replaced legacy software, today’s AI-infused startups directly address human labor — the single largest line item in most organizations.

The implication for founders and investors is huge: the total addressable market (TAM) expands from software spending into parts of the enterprise P&L previously untouched by software. In practical terms, this explains why early AI businesses can attain high ARR faster and why investors are willing to pay premiums for teams that can credibly attack $10B or even $100B markets. And because recruiting is now a central mechanism for capturing that potential, AI in recruiting becomes a core operating lever for startups looking to scale effectively.

Portfolio check-ins: who’s winning and why

Every investor on the panel had optimistic stories and caveats. Astasia highlighted companies like Synthesia and ADA (customer service automation) as examples of startups that have turned product-market fit into durable enterprise traction. These companies are landing and expanding within Fortune 100 customers — a classic signal of strong net revenue retention and product-market fit.

Matt pointed to the pace of acceleration: many top AI companies reach milestones like $5M ARR in fewer months compared with historical SaaS leaders. That rapid growth can justify higher early valuations — provided retention and gross margins hold up as they scale.

D.A. emphasized that while health-care and biotech innovation are promising, they operate on different timelines. Iterative Health’s FDA approval and go-to-market success shows the late-stage potential when life sciences companies reach commercialization, but earlier-stage drug discovery ventures still need patient capital over years — a tougher recruitment and hiring proposition in the current climate.

AI in recruiting: why every founder should pay attention (and how to act)

This is the section I want founders and hiring leaders to read carefully. When I introduce the phrase AI in recruiting here, I'm not just referencing novelty features. I'm pointing to a fundamental operational change in how teams are sourced, screened, engaged and onboarded — and how startups allocate precious headcount dollars.

Below I’ll unpack five ways AI in recruiting is changing the hiring game, with practical tips and trade-offs that mirror the broader investment and market signals we discussed on the show.

1) Sourcing and outreach become programmatic

Traditional sourcing — job boards, referrals, outreach — is being augmented by models that can read job descriptions, match candidate profiles, and personalize outreach at scale. AI in recruiting tools can comb LinkedIn, GitHub, and open-source contributions, then draft tailored messages that increase response rates. For founders, that means a smaller initial sourcing team can identify more high-quality leads.

Actionable tip: adopt AI-assisted sourcing to 2x your outreach productivity in the earliest weeks of hiring. Track conversion rates from message → interview to ensure quality remains high.

2) Screening and evaluation can be more objective — if you build it right

Bias and inconsistencies have always plagued hiring. AI in recruiting has the potential to standardize initial screens — for example, coding assessments, take-home tasks, and structured interview guides produced from job-specific competency frameworks. The caveat: models must be trained and audited for fairness.

Actionable tip: use AI to standardize first-pass evaluations, but ensure a human-led rubric and regular bias audits.

3) Candidate experience and communication scale

Candidates expect prompt, useful responses. AI chat assistants can answer FAQ-style questions, schedule interviews, and send personalized feedback. This improves offer acceptance rates and reduces recruiter time spent on basic coordination.

Actionable tip: deploy AI-driven messaging to keep candidate engagement high without sacrificing authenticity.

4) Better onboarding and role ramp plans

Onboarding drives retention. AI in recruiting extends into onboarding by generating role-specific ramp plans, learning paths, and meta-learning content — meaning the same small founding team can bring new hires to productivity faster.

Actionable tip: invest in AI-curated onboarding sequences that map role deliverables over 30/60/90 days and include measurable outcomes.

5) Strategic workforce planning — predicting needs before they appear

Beyond filling roles today, AI models can help forecast hiring needs tied to revenue milestones, product launches, and GTM plans. Accurate forecasting helps you decide whether to hire a UX designer now or contract for design work until revenue supports a headcount.

Actionable tip: tie AI-driven hiring forecasts to your financial model so every prospective hire is evaluated against impact on ARR and runway.

Gross margins, inference costs and the "subsidy" phase of AI

We spent a fair amount of time on cost structure. An important reality is that many AI products — particularly those that rely heavily on third-party foundation models — operate at lower gross margins today than traditional SaaS companies. Several AI app companies have reported gross margins closer to 50% than the typical SaaS 70-80% range. Why? Inference costs (GPU time, model hosting) and the fact that many startups are subsidizing usage while they build demand and distribution.

This phase looks familiar: it’s comparable to the ride-hailing era when unit economics were subsidized by investors to accelerate market adoption. Today, venture capital funds and strategic cloud purchases are underwriting the early inference infrastructure build-out, which in turn enables startups to scale capability quickly. The risk is a mismatch between supply and demand: if the market doesn’t adopt fast enough, cost pressures will surface.

For founders: model your unit economics at different inference price points and understand the sensitivity of gross margin to model improvements, caching, and distillation strategies. In other words, have a plan to improve margins: moving to distilled or quantized models, operating in customer VPCs, or negotiating committed cloud/GPU discounts.

Deployment patterns: bring-your-own-cloud, VPCs and on-prem options

Astasia highlighted a strategic deployment model that many infrastructure players are adopting: bring-your-own-cloud or running in the customer’s VPC. This can materially improve gross margins for vendors who can charge premium pricing for data locality, compliance and lower inference costs.

There remains a multi-modal future: some customers will prefer fully hosted SaaS models, others will demand on-prem or isolated cloud deployments for sensitivity or performance. For founders, offering flexible deployment options can be both a differentiation and a margin lever — but it requires product maturity and a savvy ops team.

Foundation models, incumbents and the acquisition landscape

OpenAI, Anthropic, and other foundation model companies creeping up the stack into applications introduces competitive dynamics. The panel agreed: incumbents and deep-stack companies will try to win horizontal productivity markets. Yet historical precedent shows that incumbents don’t always win — Snowflake and Databricks rose in spite of cloud providers building competing services.

What this means for startups: move quickly, own the customer relationship, and differentiate by vertical expertise and workflow integrations. Where horizontal concentration happens, it often creates acquisition pathways for innovative startups. If your startup plays in the productivity stack, think aggressively about defensibility (data, workflows, integrations) and potential strategic purchasers.

Valuation mania and the OpenAI case study

We spent some time debating whether companies like OpenAI at multi-hundred billion-dollar valuations are reasonable. The answer is nuanced: if OpenAI achieves $10B ARR, its current multiples may appear reasonable versus traditional software growth companies. But such growth requires both massive adoption and sustained pricing power.

From an operational standpoint, these valuations have downstream consequences for hiring and retention. Big valuations create a gravitational pull for talent and capital, which in turn makes recruiting more competitive. For early-stage founders competing for top engineers, leveraging AI in recruiting effectively can be the difference between winning hires and losing them to the largest players.

Founding team size: the rise of solo founders and lean early teams

Our conversation also covered a trend many of you have noticed: more solo founders and smaller founding teams at inception. The reasons are practical: the combination of commoditized infrastructure, AI-assisted development tools, and an increase in open-source components enables smaller teams to deliver meaningful product iterations quickly.

Still, there are trade-offs. Matt and Astasia both emphasized that while solo founders can be productive early, having a co-founder can provide emotional resilience, domain diversity, and complementary skills. It’s common for solo founders to add a co-founder early (technical or GTM) once the company proves initial traction.

How does AI in recruiting change this dynamic? Tools can offset the absence of a co-founder by making hiring, outreach, and market discovery more efficient. But the human partnership remains valuable for scaling culture and tackling ambiguous decisions — so hiring remains a strategic priority.

Hiring patterns and burn dynamics: build vs. buy people

One of the most practical themes was how hiring patterns are shifting: instead of the "raise then hire fast" playbook, many startups are designing capital-efficient growth models. AI-native tools give early teams leverage — enabling engineers to ship faster and enabling lean GTM teams to punch above their weight.

At later stages, however, companies still need to scale people: sales, customer success, ops, and product require bodies. That’s why many founders redirect the savings from smaller early teams into marketing, community building and GTM initiatives that accelerate ARR. The result: lower burn early, more focused spending on distribution, and an eventual scaling of headcount that aligns with revenue rather than vanity growth.

Outsourcing and global competition: biotech’s virtual companies and the China story

D.A. highlighted an important structural shift in biotech: the rise of virtual companies that rely on CROs and flexible service providers. This model reduces full-time headcount requirements and accelerates timelines for certain experiments and trials. Chinese startups, in some cases supported by government subsidies, have shown the speed and cost-effectiveness of this approach — executing experiments and development programs in months compared to years elsewhere.

For founders in life sciences, outsourcing strategically to CROs enables leaner core teams and better capital efficiency. But it also raises geopolitical and strategic questions about supply chains and national security when drugs or APIs are produced abroad. This is not just a hiring or recruiting problem; it’s a policy and operational challenge that overlaps with talent location and workforce planning.

Healthcare, AI and the dream of a virtual doctor

D.A. articulated one of the most compelling long-term visions: AI as the mechanism for democratizing medical knowledge. While simulating whole human biology for drug discovery may be many years away, the prospect of a "virtual doctor" — a high-quality, accessible medical knowledge system in everyone’s pocket — is closer.

Think of this as two parallel innovations: (1) the clinical and biological advances that will transform biotech over years, and (2) the immediate gains from conversational AI, voice interfaces, and decision support that can improve diagnosis, patient navigation, and clinician burnout today. Companies building conversational medical assistants or voice-based clinical documentation tools are examples where AI in recruiting matters tangibly: hiring clinical product managers, compliance experts, and ML engineers who can responsibly productize medical LLMs is a specialized hiring challenge.

Fertility tech and democratizing reproductive medicine

We closed the discussion on a hopeful note. Two portfolio companies, BillionToOne and Conceivable, exemplify how technology can improve the reproductive medicine stack. BillionToOne offers non-invasive prenatal testing that reduces reliance on invasive procedures like amniocentesis. Conceivable is automating IVF with robotics to reduce variability and cost.

Automation in embryology could dramatically lower costs, increase consistency, and expand access — turning IVF from an expensive boutique procedure into something far more widely available. The ripple effects would be large: fewer medical complications, more predictable outcomes, and possibly a shift in demographic trends in countries where infertility treatment is a meaningful factor.

Exits, M&A and the back half of the year

Both Astasia and Matt saw green shoots for exits: a resurgence of M&A, some notable acquisitions (Weights & Biases by CoreWeave, Superhuman by Grammarly) and IPO readiness in a slate of private companies. For founders, this meant a reminder: building to sale is valid, but building to IPO requires close attention to metrics, margins and credible scale.

What this implies for hiring is straightforward: if your path to liquidity is acquisition, focus on building strategic capabilities that make you an attractive target (vertical expertise, customer contracts, sticky data). If your path is IPO, you must plan for the scale of organizational infrastructure that public companies require, which means early investments in finance, security, legal and HR.

Practical advice for founders on hiring, fundraising and product

Across the conversation, several practical rules repeated themselves. I want to summarize them here as tactical advice you can act on this quarter:

1. Model runway carefully: Only raise what you need for a clear milestone. Avoid the trap of over-raising and giving up equity too early.
2. Use AI in recruiting sensibly: Adopt tools for sourcing and screening, but pair them with human-led rubrics to reduce bias and preserve culture.
3. Prioritize marketing and community early: With lower distribution friction, founders must actively build awareness to break through the noise.
4. Plan for inference economics: Model gross margin sensitivity and have a transition plan to improve margins (distillation, caching, VPC deployment).
5. Be thoughtful about team composition: One founder can start, but consider adding complementary skills early; also plan to hire focused GTM hires once product-market fit is clear.
6. Think global: For biotech and life sciences, partnerships with CROs and international teams can accelerate timelines and disciplines.

How investors are thinking about the next 6–12 months

The panel’s investment view was pragmatic: be selective, play the field, and recognize the new regime's dynamics. Matt emphasized discipline — you have to play on the field where the puck is, but with rigor. Astasia emphasized the opportunity in attacking labor budgets and building products that can expand enterprise margins. D.A. emphasized long-term conviction in health care and biotech while acknowledging the short-term capital constraints.

From a recruiting and talent perspective, investors are looking for founders who can build teams that are simultaneously efficient and deeply capable — a combination increasingly enabled by AI in recruiting tools and a focus on community-driven hiring.

Closing thoughts: the interplay of AI in recruiting and strategy

We closed the episode optimistic about a strong back half of the year. Why? Because innovation frontiers are expanding: AI is creating larger addressable markets, infrastructure companies and AI apps are scaling, and exits are resuming. At the same time, biotech and life sciences face a timing puzzle that requires patience and specialized capital.

For founders and hiring leaders, the most actionable takeaway is clear: adopt AI in recruiting as a strategic capability. Use it to accelerate sourcing, reduce bias, structure onboarding and forecast hiring needs — but pair these tools with disciplined long-term people planning. The startups that balance technical capability, product-market fit, and thoughtful hiring will capture disproportionate value in this new landscape.

Further reading and next steps

If you want to go deeper, here are a few follow-up actions inspired by the conversation:

• Audit your current hiring funnel and identify where AI in recruiting could reduce time-to-hire and improve conversion rates.
• Model your unit economics at multiple inference price points and identify the break-even threshold for your product.
• Build a 6–12 month talent roadmap tied to ARR milestones — prioritizing GTM hires that move the needle on revenue.
• For biotech founders: consider virtual company models and CRO partnerships to conserve capital and accelerate experiments.
• For early-stage founders: invest in marketing and community early. Product excellence needs distribution to become durable.

Final note

The era we described is both exhilarating and fragile. There’s a supply-side surge — massive investment in infrastructure, models and product — and demand must follow. Investors and founders are placing different bets: some are aggressive buyers at high valuations, others are patiently waiting for better entry points. No matter your stance, integrating AI in recruiting thoughtfully will be a key operational advantage. It helps you compete for scarce talent, scale efficiently, and convert your product advantages into lasting company value.

Want to continue the conversation? Reach out, and let’s talk about how to apply these lessons to your own hiring and growth strategy.