Garrett Lord’s first big sale was convincing a Fortune 500 company to post a job on his fledgling student network. A decade later, his most important customers aren’t hiring graduates; they’re hiring the graduates themselves to train the models that might replace them. Handshake AI, a unit launched in 2024, has turned the company’s network of 18 million students and alumni into a high-skill workforce for frontier AI labs, scaling to an estimated $100 million in annualized revenue within eight months [AI Native GTM Substack, ~2024]. For procurement officers at those labs, the pitch is straightforward: verified expertise, at scale, without the overhead of building a freelance marketplace from scratch.
The Wedge Was Already Built
Handshake’s core business is a recruiting platform used by every Fortune 500 company to connect with early talent from over 1,500 universities [Handshake LinkedIn, unknown]. That network, particularly its 500,000-plus PhDs across 200 specialties, represented a latent asset. When AI labs like OpenAI and Anthropic began scrambling for domain experts to validate, annotate, and prompt-train their most advanced models, they faced a quality and verification problem. Handshake’s existing university partnerships and verified academic credentials provided a ready-made solution [Handshake Blog, ~2024]. The AI unit didn’t need to recruit a new crowd; it simply needed to redirect a fraction of an existing one toward a new, higher-margin task.
Traction That Outpaced the Core Business
The speed of the unit’s growth is the most compelling data point. According to one analysis, Handshake AI went from zero to an estimated $50 million in annualized revenue in four months, hitting the $100 million mark two quarters later [AI Native GTM Substack, ~2024]. For context, the parent company’s entire revenue was reported at $172 million in 2024 [GetLatka, unknown]. This suggests the AI services segment, in its first year, became the company’s primary growth engine and may have already eclipsed the decade-old recruiting business in sheer momentum. The company claims it has paid out over $100 million to more than 100,000 fellows working on AI training tasks [Handshake website, ~2025].
| Metric | Core Recruiting Business | Handshake AI Unit |
|---|---|---|
| Launch Year | 2014 | 2024 |
| Network Size | 18M+ students/alumni | 500K+ PhDs (subset) |
| Key Customers | Fortune 500 employers | OpenAI, Anthropic [AI Native GTM Substack, ~2024] |
| Reported 2024 Revenue | $172M (total company) [GetLatka, unknown] | $100M annualized (unit) [AI Native GTM Substack, ~2024] |
| Primary Use Case | Job recruitment & placement | Model validation, annotation, prompting |
The Founder-Mode Pivot
CEO Garrett Lord, a Forbes 30 Under 30 alum who famously cold-emailed his way into a Palantir internship from Michigan Tech, has enforced what he calls a “founder-mode” culture [Forbes, 2022]. This intensity was directed at the AI opportunity. Lord personally recruited technical leadership for the new unit, including Sahil Bhaiwala, formerly General Manager and Director of Product & AI at Scale AI, a direct competitor in the data labeling space [Handshake Blog, ~2025]. This hire signals a deliberate move to compete on product and systems, not just network access. The founding team,Lord, Ben Christensen, and Scott Ringwelski,are all Michigan Tech computer science alumni who have built the company through multiple funding rounds totaling $434 million [TechCrunch, 2021]. Their bet is that deep, vertical integration between their network and an AI-specific tech stack will create a moat.
Where the Model Could Stumble
Explosive growth from a standing start is impressive, but it surfaces questions about sustainability and focus. The model’s advantages are also its potential constraints.
- Capacity ceilings. The network is large, but the subset of PhDs and domain experts willing and able to perform high-quality AI training work is finite. Scaling revenue tenfold again might require expanding beyond the current academic pool, diluting the verified-expertise differentiator.
- Internal competition. Handshake must balance its legacy mission,helping students find full-time jobs,with incentivizing them toward lucrative, but potentially transient, gig work for AI labs. A significant bifurcation could strain university relationships.
- Competitive response. The success has drawn attention. Specialized AI training startups like Surge AI and talent platforms like Mercor are chasing similar contracts. Larger players like Scale AI, with whom Handshake now competes directly for enterprise deals, have deeper war chests and established ML infrastructure.
The most credible risk is that the AI unit’s growth becomes a distraction that commoditizes the core network asset. Handshake’s answer appears to be a focused, semi-autonomous unit with dedicated leadership, betting that the two businesses can coexist by serving fundamentally different customers with the same underlying resource.
The Next Twelve Months
The immediate milestone is proving the $100 million run-rate is not a flash in the pan but a repeatable baseline. Watch for two signals: whether Handshake AI begins disclosing named enterprise customers beyond the initial AI labs, and if the parent company seeks a dedicated funding round for the AI unit to fuel further product development and international expansion. The hire from Scale AI suggests a roadmap that includes more sophisticated annotation tools and managed services, moving up the value chain from labor marketplace to full-stack training partner.
The ideal customer profile here is clear: the head of AI research or data operations at a well-funded lab or large enterprise building proprietary models. They have a budget for quality assurance that dwarfs typical crowd-sourcing costs, and their primary constraint is access to reliable, scalable subject-matter expertise, not price. For them, Handshake AI is selling risk reduction.
The realistic competitive set includes pure-play AI training platforms, general freelance marketplaces trying to move upmarket, and the internal sourcing teams at the labs themselves. Handshake’s defensibility rests on the depth of its university integrations,a barrier built over ten years and $434 million of venture capital,not on any proprietary model. In a market hungry for intelligence, they’ve found a way to monetize the one asset they had in abundance: verified human expertise.
Sources
- [AI Native GTM Substack, ~2024] How Handshake reinvented itself for the AI era and built a $100M segment in 8 months | https://ainativegtm.substack.com/p/how-handshake-reinvented-itself-for
- [Handshake Blog, ~2024] Introducing Handshake AI | https://joinhandshake.com/blog/our-team/introducing-handshake-ai/
- [Handshake LinkedIn, unknown] Company post on Fortune 500 employer reach | https://www.linkedin.com/company/joinhandshake/
- [Handshake website, ~2025] Handshake AI program details and payout figures | https://joinhandshake.com/ai
- [GetLatka, unknown] Handshake revenue estimate for 2024 | https://getlatka.com/
- [Forbes, 2022] Garrett Lord profile and Forbes 30 Under 30 recognition | https://www.forbes.com/sites/kristinstoller/2021/05/12/handshake-a-job-search-platform-for-college-students-valued-at-15-billion-after-new-funding-round/
- [Handshake Blog, ~2025] Announcement of Sahil Bhaiwala joining Handshake AI | https://joinhandshake.com/blog/our-team/introducing-handshake-ai/
- [TechCrunch, 2021] Handshake raises $80M at a $1.5B+ valuation | https://techcrunch.com/2021/05/12/handshake-raises-80m-at-a-1-5b-valuation-as-its-diversity-focused-recruitment-network-for-grads-passes-18m-users/