Tenex: The 10x Shift in AI Engineering Compensation

“There’s like a moment where you get smacked in the face by how profound this technology can be, if harnessed in the right way.” This sentiment, articulated by Alex Lieberman, co-founder of Tenex, encapsulates the core insight driving his and Arman Hezarkhani’s venture. In a recent discussion, Lieberman and Hezarkhani, founders of the AI-first software consultancy Tenex, unpacked how they are fundamentally rethinking knowledge work compensation, specifically for AI engineers, in an era where artificial intelligence can dramatically amplify productivity. Their conversation offered a sharp analysis for founders, VCs, and AI professionals grappling with the economic shifts brought about by generative AI.

The genesis of Tenex stems from an unexpected crucible. Lieberman, an early investor in Hezarkhani’s previous company, Parthian, witnessed firsthand a profound transformation. Faced with a 90% downsizing of his engineering team, Hezarkhani was compelled to re-architect Parthian’s entire product and engineering process to be AI-first. The astonishing outcome was a 10x increase in production-ready software output, despite the massive reduction in headcount. This counterintuitive result illuminated a critical flaw in traditional compensation models: hourly billing perversely incentivizes slower work, even as AI tools enable unprecedented speed and efficiency. Lieberman, initially skeptical, became a believer after Hezarkhani detailed the specific impact large language models (LLMs) were having on engineering as a form of knowledge work.

Tenex’s disruptive economic model addresses this misalignment by compensating engineers based on "story points"—units of completed, quality output—rather than hours worked. This system directly incentivizes engineers to embrace every new AI tool, optimize their workflows, and maximize their throughput. Hezarkhani openly shared that Tenex anticipates multiple engineers will earn over $1 million in cash compensation next year purely from story point earnings. This radical approach attracts top-tier talent, but also raises questions about system integrity. Tenex mitigates the risk of "gaming" the system by hiring for two key profiles: engineers who are "long-term selfish" and understand that inflating story points will ultimately damage client relationships, and those who possess a genuine passion for writing code and collaborating with intelligent individuals. Furthermore, technical strategists, whose incentives are tied to client retention (Net Revenue Retention or NRR), serve as the final quality gate before any engineering plan reaches a client.

The results speak for themselves, demonstrating AI's transformative power on project timelines and sales motions. In one instance, Tenex developed a sophisticated computer vision system for retail cameras—providing heat maps, queue detection, shelf stocking analysis, and theft detection—creating early prototypes in a mere two weeks for work that traditionally took quarters. Another project saw them build Snapback Sports' mobile trivia app in just one month, which subsequently climbed to 20th globally on the App Store. The agility AI affords even reshapes client acquisition: an engineer once spent four hours building a working prototype of a fitness influencer’s AI health coach app after the prospect had initially declined their services, immediately catapulting Tenex to the top of their vendor list.

Despite being an AI-first company, Tenex’s primary constraint is human capital. The challenge lies in finding and hiring enough exceptional engineers quickly enough, and then seamlessly integrating them into processes that maintain delivery quality as the company scales. Technology, while foundational, acts as an enabler; the true bottleneck remains the acquisition and effective deployment of top human talent.

The interview process at Tenex reflects this commitment to exceptionalism. While many firms might abandon take-home assessments in the age of AI, Tenex makes theirs "unreasonably difficult." Roughly half of candidates don't even respond, but those who complete the challenge demonstrate the caliber required. The interview process is remarkably short: two calls before the take-home, review, then one or two final meetings, completable in as little as a week. A signature question posed to candidates is, "If you had infinite resources to build an AI that could replace either of us on this call, what would be the first major bottleneck?" The sophisticated answer, according to Tenex, is not merely "model intelligence" or "context length," but rather "controlling entropy," the accumulating error rate that can derail autonomous agents over time. This focus underscores Tenex's deep understanding that even with advanced AI, human oversight and strategic thinking remain paramount.