Deloitte Picks Two‑Person Startup for First Enterprise AI Deal

A Two-Person Startup Bets One Intern Can Replace a Sales Floor

The job listing fits on a phone screen: "Technical GTM Intern," remote from India, ₹40,000 to ₹100,000 a month, posted by a two-person Y Combinator F25 company called Nivara. It tells you a lot about where early-stage space and enterprise-AI startups think sales headcount is going.

Nivara, founded in 2025 by ex-Uber engineers in San Francisco, posted the role on its YC jobs page. The startup describes itself as "a small applied AI team working on enterprise finance and operations," with agents embedded in finance workflows such as bank-line reconciliation, invoice matching, and exception review. The home page frames the pitch bluntly: "This work wasn't automatable two years ago. It is now."

A remote intern handling GTM engineering for a company whose own agents already automate back offices is a deliberate test. The role blends traditional sales operations with the AI automation Nivara ships to customers: pipeline tooling, prospect research, outbound sequencing, and the integration plumbing that, until recently, sat inside a RevOps org chart. That same work, at US salaries, costs roughly $127,500 a year at the median, according to Prospeo's January 2026 analysis of more than 1,000 GTM engineering postings. Nivara's intern slot, priced for an Indian remote hire at a fraction of US GTM-engineer pay, suggests the startup is testing whether one human alongside the very automation Nivara sells can replace the SDR-and-ops stack that used to staff an early-stage sales motion.

The intern hire is small in absolute terms and large in what it signals about Nivara's sequencing. The company's product page lays out the rule: "We take on work that's internal and measurable before anything customer-facing. That's a rule about sequencing, not a limit on scope." A remote GTM engineering intern is internal and measurable. If Nivara's pitch holds, the role is automatable in months, not years.

The Category That Didn't Exist Three Years Ago

The math behind Nivara's intern-first GTM bet isn't a quirk of one Y Combinator graduate. It's the visible edge of a category that didn't exist three years ago and now hires faster than almost any other slice of tech.

Clay coined the term "GTM engineer" in 2023, and the title didn't appear in job postings at scale until 2024. By mid-2025, recruiters tracked roughly 1,400 open GTM engineer roles; by January 2026, that number had blown past 3,000 on LinkedIn alone. GTM engineering and related RevOps postings climbed 205% year-over-year, according to Prospeo's January 2026 index — roughly 13 times faster than the broader tech labor market.

What the role actually does explains the velocity. A single GTM engineer can source pipeline for a sales team of six to ten, and Clay's Verkada case study showed automating roughly 80% of SDR workflows let reps book four times as many meetings per month — 80 to 100 each. Rootly reported a 69% jump in scheduled meetings after similar work. Only 1.4% of GTM engineering postings mention cold calling. SQL and Python each appear in 38% of them. Most GTM stacks already run five to ten tools, and someone has to make them talk to each other; AI now lets one person do this at the scale of a department.

"The aperture for how you build your go-to-market or revenue engine and the decisions you make to get there have never had more unique and specific pathways depending on your company." — GTMfund partner Paul Irving, speaking to TechCrunch in January 2026

Remote work accelerated the shift. Roughly 45% of people carrying "GTM Engineer" in their title already work as agencies or consultants rather than full-time staff. That is exactly the labor shape an early space startup needs: one or two embedded builders wiring AI to a CRM and an outreach stack, not a five-person sales org before the first contract closes.

For space founders watching their peer cohort adopt the same playbook, the implicit message from Nivara's intern-first structure is that the same plumbing will soon wrap around their own pipelines. GTMfund has built its thesis explicitly around "distribution as the final moat in the AI era," and early backers don't want to see half a seed round spent on a sales team before the product ships.

If the experiment works, expect the job listings to follow the money. Roughly 9 out of 10 responsibilities in GTM engineering postings also appear in RevOps postings, so the boundary stays blurry. But the pay bands and headcount mix signal where early space startups will land next. Hire one GTM engineer with AI in their toolchain, not five SDRs with dialers.

Investors, Founders, and a Big Four Firm All Pull the Same Direction

The reactions framing Nivara's experiment as a signal, not an isolated stunt, are accumulating from three directions: venture investors repositioning around AI-native go-to-market theses, peer founders quietly copying the playbook, and enterprise buyers normalizing the pattern through procurement decisions.

The investor signal is concrete. TechCrunch's January 2026 coverage of GTMfund, including Paul Irving's appearance on the Build Mode podcast, frames the thesis in plain terms: the traditional go-to-market playbook "may have worked in the days of traditional enterprise SaaS but won't cut it in the crowded, AI-driven startup era of 2025." The firms underwriting that thesis have begun to swap it for an AI-first model, and Nivara's intern-plus-automation structure is the kind of unit economics they want to underwrite.

The peer signal is structural, not anecdotal. TechCrunch reporting on India's startup rules for deep tech, published in February 2026, captured the broader capital reallocation behind the shift: "India is adjusting startup rules, and mobilizing public capital, hoping to help more of them make it to commercial products," with deep-tech time horizons stretched from 10 to 20 years. The same principle applies directly to early space startups weighing sales hires: patient capital forcing founder efficiency rather than headcount expansion. Roughly 45% of GTM engineering titles already sit with agencies or consultants, a remote-talent pattern Prospeo tracked in its January 2026 index, which "explicitly opens the door to the kind of remote intern role Nivara posted."

The enterprise-partner signal carries the most weight because it converts an experiment into a procurement decision. Deloitte's selection of Nivara as its first Enterprise AI partner puts a Big Four firm's reputation behind the thesis that AI can own workflows in regulated industries. The CIO.com analysis put the inflection in commercial terms: AI shifts from demo to deployment "when CIOs wire it into the whole GTM flow, turning customer signals into faster deals, stronger pipelines and real revenue wins."

The peer cohort moving fastest is the one with the least room to hire. Early-stage space startups facing 2026's funding discipline don't have a sales-expansion lever, so they're testing Nivara's substitute. Microsoft, AWS, and other infrastructure incumbents are reorganizing around AI-native revenue operations, which raises the floor for what a two-person space startup is expected to automate. Sam Altman's reported attempt to acquire a rocket company to challenge SpaceX, while not a GTM story itself, signals that the orbital-AI race is now considered strategically adjacent to commercial revenue machinery, not a separate procurement line.

The open question isn't whether the pattern replicates — it already is — but whether the intern-plus-AI structure scales past Nivara's first cohort. Watch the next Y Combinator batch announcements: if GTM engineering intern postings become a standard line item alongside founding engineer listings, Nivara's pilot graduates from experiment to template.

Working in frontier tech? Zero G Talent tracks the openings: see every open ASML role, browse frontier tech jobs, openings at Stripe, and the people building the field.