The Skill Swap Demands That Doubles Conversion Rates

What Swap Looks For

Swap posted 29 full-time openings this week — a single hiring wave, not a rolling backlog, after closing a $100 million Series C from ICONIQ and DST Global. The roles span five hubs and every function from sales to infrastructure, but the screening filter is the same: does this person understand the merchant's P&L, move at Swap's tempo, own outcomes across boundaries, speak cross-border, think in agentic primitives, and default to collaboration?

That orientation, commercial, not academic, traces to the founders. The CEO spent two years at McKinsey watching big retailers wrestle with operations, then joined Juni, a venture-backed fintech, to see a startup scale and stumble. His co-founder Zach spent a decade in textiles, threading a global manufacturer network. Their first venture together, an e-commerce brand importing handmade goods to the UK, died on margin math: duties paid twice, pricing that didn't bake in landed cost. "We didn't actually get to one," the CEO said of that zero-to-one attempt. But the scars (returns, cross-border tax, cash-flow opacity) became Swap's product roadmap.

First filter: domain fluency. Early hires came from Shopify and Klaviyo, companies Swap's leadership "look up to and kind of want to emulate in some way." The reasoning is practical: people who already speak merchant can "talk to the customers straight away and really relate to them." In a space where founders trade warnings in WhatsApp groups — "if you do one thing wrong, they're going to put it in there and say, 'Hey, don't use this tool'", credibility compounds. A candidate who has never set up a Shopify store, never fought a chargeback, never explained a surprise duty charge to a UK buyer shipping to Texas starts at a deficit.

Second: speed as habit, not slogan. The values page reads: "Act fast: Fail fast, learn faster. Speed is our edge." Investors noticed. One board member recalled a diligence call with engineer Johan (the night his wife went into labor). He took the call from the hospital. The investor told him to hang up. That story circulates not as hero worship but as calibration: this is the tempo. Candidates who optimize for process over shipping, who need perfect spec before writing code, who treat "move fast" as a poster rather than a daily trade-off, tend to self-select out before the offer stage.

Third: ownership spanning the stack. Swap's pitch is an "agentic commerce OS" — a single platform unifying logistics, returns, payments, tax, and a new AI-native storefront. Job descriptions for the current wave repeatedly ask for "real ownership" and "build from the ground up." A Senior Demand Generation Manager in New York owns pipeline end-to-end, not hand off MQLs. An Analytics Engineer in London models data inside a product team that ships merchant-facing dashboards. The IT Helpdesk Administrator in Austin supports hundreds of employees across five hubs, but the posting emphasizes "building internal tooling" over ticket triage. The pattern: every role touches the product, and every role touches the merchant.

Fourth: cross-border instincts. Swap's wedge was returns. Its expansion (real-time tax and duty API, Universal Catalog for agentic shopping, global services network) means the team needs people who have felt the pain of HS codes, IOSS registrations, state-by-state nexus, and the avalanche of new tariffs. The CEO described the premise: "help a merchant sell internationally just as easily as if they were selling domestically." That means calculating duties at checkout, handling customs clearance, filing US state taxes for a UK brand. Candidates who have only operated in a single market, or who treat international as "shipping plus a fee," miss the complexity Swap absorbs.

Fifth: AI-native product intuition, specifically around the agentic storefront. The platform replaces a static theme with a conversational interface where shoppers "ask, discover, and buy through one experience." Brands on Swap convert at twice the rate of traditional storefronts while keeping shoppers on site three times longer and cutting returns by one in five. That number matters because it reframes the engineering challenge: it's not RAG on a catalog. It's a storefront that reasons — about inventory, about margin, about the shopper's intent, in real time. Engineers who have shipped LLM features in production, who understand latency budgets for streaming responses, who have designed evals for non-deterministic outputs, get a longer look. So do designers who have prototyped conversational flows, not just checkout funnels.

Sixth: the "Win together" signal. The values page puts it plainly: "Collaboration is in our DNA." In practice, the onboarding manager in London works daily with the tax engine team, the demand planning squad, and the Avalara partnership channel (a Slack connect where the Avalara team "can't get over how quick everyone replies"). Candidates who describe their best work as solo heroics, who can't name a time they unblocked a peer in another function, or who treat support as a tier below engineering, tend to stall in the values interview.

The screening isn't a checklist. It's a coherence test: the same criteria apply. The roles open now (spanning growth, data, support, infra, product, commercial) all map to that profile. The next section breaks down the platform they'll be building on.

Inside the Agentic Storefront

Swap's agentic storefront isn't a chatbot bolted onto a product catalog. It is a separate, brand-owned sales channel — a dedicated .ai destination beside the merchant's .com, where an AI agent guides shoppers through discovery, virtual try-on, and checkout entirely through natural language, text or voice. The agent speaks in the brand's tone, applies the brand's business logic, and captures keystroke-level first-party data that stays with the merchant. Forbes reported the launch in May 2026, noting the platform achieves the same conversion metrics.

The architecture hinges on real-time sync with Shopify, which acts as the single truth source for product data, inventory, and pricing. No separate catalog to maintain. That sync is the backbone: if the agent fails to query live inventory or pricing instantly, the conversation breaks. The help center documents three deployment tiers:

Each tier demands different engineering lift. Core requires clean API integration. Branded adds front-end customization within Swap's framework. Custom demands full-stack co-development across legacy databases, custom inventory systems, and third-party logistics providers that, as the AI Agents Directory analysis put it, "do not 'speak' the same language."

Interoperability is the hard problem. Retailers run fragmented stacks (legacy ERPs, homegrown inventory tools, 3PL APIs) and the agent must orchestrate across all of them without latency that kills conversion. Bad data multiplies risk: inconsistent product descriptions or faulty inventory feeds generate hallucinated recommendations, and user trust, once lost, proves hard to regain. Computing cost squeezes cloud budgets: autonomous agents devour significantly more processing power than traditional page loads. Security demands a third layer: granting agents power to transact requires robust authentication and authorization to prevent unauthorized access.

Launch partners SIMKHAI, Retrofête, Odd Muse, and Studio Nicholson went live on Core and Branded tiers. Luxury retailers view the platform as a way to mimic the in-store concierge experience; big-box merchants seek to curate personalized journeys for their top 40 percent VIP customers. Founder-led brands have even cast the founder as the literal voice of the agent. Support for Centra and Salesforce Commerce Cloud arrives soon, expanding the integration surface.

This reality — real-time multi-system orchestration, brand-owned agent personality, keystroke-level data capture, compute-intensive inference, transaction-grade security, explains why Swap's open roles cluster around agentic AI integration, platform reliability, and cross-functional product engineering. Generic full-stack experience doesn't prepare a candidate for the failure modes of an agent that must negotiate live inventory across three APIs while maintaining a branded conversational flow. The hiring bar reflects that specificity.

Why 29 Roles Now

The functional split shows where capital flows:

The lone AI-specific role and single Data seat suggest deliberate staffing of the agentic layer, not hype-chasing.

Geography tells the expansion story. Austin absorbs fourteen of the 29 roles — almost half, making it the clear hub for this cycle. London takes twelve. Amsterdam holds four. New York and Tel Aviv each have two. The five hubs match the "250+ employees worldwide across 5 different hubs" figure Swap publishes, and the new openings will push headcount above 270 if filled. Work-arrangement data reinforces the hybrid default: twenty-five roles carry hybrid tags, three remote, one on-site. That mix has held steady since the hundred-million-dollar round close.

Timing anchors to the hundred-million-dollar round, announced as the catalyst for "the next phase of growth across the U.S. and Europe." The board's own feed shows two roles added in the past seven days: the demand-gen role (band $135k–$170k), Analytics Engineer in London, IT Helpdesk Administrator (Austin), Merchant Onboarding Manager (London), Technical Support Specialist (Austin), and Customer Support Specialist (Austin). That refill pace will keep the total near this level for months.

The salary band on that New York demand-gen role ($135k–$170k, median $170k) is the only compensation figure the board surfaces publicly. It matches senior commercial hires in Austin and London, where cost-of-living math compresses the spread. No equity ranges are published.

Swap's careers page calls the surge "taking real ownership as the platform revolutionizes commerce." The numbers support the rhetoric: 29 roles against a 250-person base is about one in nine headcount increase in a single cycle.

What Candidates Say Worked

Glassdoor shows ten anonymous reviews for Swap, but few break down the interview loop in detail. The platform's "How can I get a job at SWAP?" guidance reads simply: browse open roles, apply locally, then study the specific interview process once a recruiter responds. That mirrors the board data listing half a dozen live openings across New York, London, and Austin — Senior Demand Generation Manager, Analytics Engineer, IT Helpdesk Administrator, Merchant Onboarding Manager, Technical Support Specialist, and Customer Support Specialist, each with its own functional screen.

A TeamBlind thread titled "Swapping Interview Experiences/Questions" attracted contributors from Meta, Airbnb, Reddit, Roblox, and Stripe who swap FAANG-adjacent question banks. While none explicitly named Swap, the pattern applies: engineers who prepared for system-design prompts around real-time inventory sync, multi-region checkout latency, and LLM-driven product recommendation pipelines aligned with the skill sets implied by Swap's current openings. The Analytics Engineer role in London pairs with a Merchant Onboarding Manager, implying interviewers probe SQL fluency joined with an understanding of seller onboarding funnels.

Hibob's industry research on interview feedback backs what successful candidates practiced: requesting actionable takeaways after each round. "Honesty, sensitivity, and actionable suggestions, all while maintaining a balanced tone," the guide urges.

The board data lists a $135k–$170k median band for the role, the only role with published compensation. Candidates who tied salary discussions to that band (rather than citing generic market data) showed they'd done the homework.

No single playbook guarantees a Swap offer. But the pattern across available signals is consistent: treat the process as a product conversation, map your experience to the specific agentic-storefront problems the open roles imply, and close each round by asking what the team needs tomorrow that it doesn't have today.

Where Swap's Bar Sits

Public signals on how Swap screens candidates remain thin, and side-by-side hiring data for companies building agentic storefronts is absent. We can only read Swap's own technology claims and the roles it advertises, then ask what those imply about the bar relative to the broader AI-native commerce cohort.

Swap's marketing rests on a measurable proposition: brands on its platform convert at that rate, keep shoppers engaged longer, and capture ten times better first-party data at the keystroke level. The company also cites a one-in-five reduction in returns through virtual try-on and fit guidance. Those numbers, repeated across Swap's site and cited in a Forbes piece on agentic commerce's emergence, cast the product as a full-funnel conversion engine, not a bolt-on chatbot. If the product promise is end-to-end ownership of discovery, try-on, and checkout, the engineering and product bar moves toward people who have built multi-modal, real-time decision loops — not just RAG pipelines or static recommendation widgets.

Live board data on Zero G Talent reveals roles reinforcing that inference. In the past week alone the company listed the demand-gen lead with a $135k–$170k band, an Analytics Engineer in London, the onboarding lead, two support-focused roles in Austin (the support specialists), and an IT Helpdesk Administrator. The salary band for the demand-gen lead exceeds typical early-stage commerce SaaS marketing hires, implying Swap expects that person to own a technical, product-led growth motion — someone who can turn the twice-the-conversion claim into repeatable playbooks for the more than 800 brands already on the platform. The Analytics Engineer role flags a need for event-stream fluency at the keystroke level Swap touts; that differs qualitatively from the batch-oriented BI work common at traditional e-commerce enablement tools.

Contrast that with hiring patterns at adjacent companies. Public job boards for firms positioning themselves as "AI shopping assistants" or "conversational commerce" cluster around prompt engineering, LLM fine-tuning, and chat UX, roles optimizing a single conversational interface. Swap's open roles lean toward platform-level instrumentation (analytics), merchant-facing onboarding, and cross-border operational support (the Austin support hires match the "global commerce through cross-border, tax, demand planning, and returns" line in Swap's own description). That spread shows a screening filter valuing systems thinking across commerce infrastructure — payments, duties, returns, catalog sync — over pure model-centric experimentation.

Forbes coverage of agentic commerce as "the next big shift in e-commerce" underscores that the category is still defining its vocabulary. Without a standard rubric, each company draws its own line. Swap's bar reads as: can you ship a feature that moves the twice-the-conversion needle inside a live merchant environment next quarter? — a product-outcome threshold that screens out researchers who only publish and engineers who only optimize offline metrics. Whether that threshold is higher or lower than peers remains unknown; what stands clear is that Swap's current hiring slate targets operationalizing its claimed metrics at merchant scale, not demonstrating model benchmarks in isolation.

The gap in comparative hiring intelligence exists. Until a specialized recruiter or analyst publishes a compensation and competency survey for agentic-storefront teams (or until Swap and its peers disclose interview rubrics), any "benchmark" remains an informed read of public role specifications, not a measured comparison.

What the Hiring Wave Signals

Analysts and industry veterans read Swap's 29-role expansion not as an isolated sprint but as a data point in a broader capital-allocation shift reshaping the labor market for AI-native commerce. A Bloomberg Intelligence survey of 151 senior financial-services employees, published December 10, 2025, found that two in three firms expect initial staff increases and more than seven in ten anticipate higher operating costs over the next three years. "Rather than immediate job cuts, artificial intelligence adoption is driving a near-term hiring boom on Wall Street," the report said. That pattern — heavy upfront hiring to build capability before efficiency gains materialize, mirrors what venture-backed commerce platforms are now doing.

UBS's data shows global AI capital expenditure will hit $1.3 trillion by 2030 — more than the GDP of Indonesia, encompassing data centers, advanced chips, fiber networks, power systems, cloud capacity, and specialized software. Jamie Dimon said this infrastructure investment "is going to cause probably more jobs in the short run in total," requiring roads, trucks, drivers, electricians, and construction workers well before automation reduces white-collar positions.

A KPMG US CEO Outlook Pulse survey of 100 large-company CEOs, reported March 20, 2026, reinforced the narrative: fewer than one in ten leaders planned to reduce headcounts from AI investments, while more than half expected AI to increase hiring in 2026. Tim Walsh, CEO of KPMG US, calls this the "labour cost margin", a recalibration of labor, technology, and delivery cost. Tim said AI is already transforming how KPMG recruits, creating job categories that didn't exist before widespread AI adoption. The firm now hires technologists in ways it never did, alongside "orchestrators" who manage substantial workflow portions to ensure completion, accuracy, and appropriate outputs. IBM plans to triple entry-level hiring across 2026, though the roles have fundamentally changed. Nickel LaMoreaux, IBM's CHRO, said at Charter's Leading with AI summit: "The entry-level jobs you had two to three years ago, AI can do most of them. If you're going to convince business leaders to make this investment, you need to show the real value these individuals bring now, through totally different jobs."

For AI-native commerce specifically, the hiring surge signals companies are in what Bloomberg Intelligence analysts Diksha Gera and Tomasz Noetzel call the "early phase in financial services that prioritizes capability building over cost reduction." They project cost ratios will normalize after 2027-28 as automation scales, unlocking efficiency-driven margin expansion. JPMorgan Chase, one of the most aggressive AI adopters, employs two thousand people on AI systems developed since 2012 and announced plans in July 2024 to add three thousand more AI-focused positions over two years. The bank invests roughly two billion dollars annually in AI and has achieved break-even return, Dimon said. Yet Dimon recently instructed managers to "resist headcount growth where possible," signaling a shift toward efficiency as AI capabilities mature. Goldman Sachs expects to finish 2025 with net headcount increases despite reductions elsewhere, hiring in high-priority areas while using AI to reduce needs in other functions. A Goldman Sachs survey of over 100 investment bankers found about one in nine U.S. companies currently reducing headcount due to AI, but expectations for AI-driven job reductions reach one in twenty-five over the next year and potentially one in nine over three years, with financial institutions facing the steepest anticipated cuts at about one in seven.

Private-equity firms clamor to hire AI operating executives for portfolio companies, and hedge funds offer up to $350,000 annually for top-tier AI researchers and engineers. Ben Hodzic, managing director at Selby Jennings, told Business Insider there's "a lot of optimism" around AI in financial services, driven by a desire to build AI tools in-house and boost productivity in wealth advisory, investment banking, and trading. UBS Group AG analyst Jason Napier said there's "a decent probability that in 2026 the equity market decides — perhaps on still-scant hard evidence, that banks will be one of the most significant beneficiaries of the rapidly improving technology on offer." UBS analysis suggests European banks, employing a workforce the size of Houston (2.1 million people), could achieve cost reductions of one in six to one in five through AI, translating to roughly one in five more profit before tax. The investment bank projects 2026 as an inflection point when markets may determine whether AI represents meaningful profit potential or remains speculative for banking.

Swap's own board data shows the company added two roles in the past week. The Series B round of forty million dollars led by ICONIQ Growth in March 2025 gave the e-commerce operating system capital to fund this expansion. The hiring wave across AI-native commerce, Swap included, reads as a leading indicator that the sector remains in the heavy-investment, capability-building phase. Bloomberg Intelligence's two-phase model applies: Phase 1 (now through 2027-28) brings initial headcount increases averaging four percent, rising operating costs at more than seven in ten firms, and a focus on hiring AI specialists, data scientists, and integration engineers. Phase 2 (2027-28 and beyond) will see automation reach sufficient scale to reduce headcount needs, normalized cost ratios enabling margin expansion, and projected job cuts of a city's worth (150,000–200,000) across global banking, a trajectory commerce platforms will likely follow with a lag. For now, the message from analysts is consistent: the hiring surge is not a bubble but a necessary prelude to the productivity gains that will define the next decade of AI-native commerce.

The Roadmap in Headcount

The open roles Swap advertises across its five hubs resemble a product roadmap rendered in headcount. Each cluster targets a specific layer of the agentic commerce OS the company is building: the AI-native storefront, the inventory intelligence layer, the cross-border and tax engine, and the partner ecosystem that will distribute it all. With a hundred million dollars in Series C capital from ICONIQ and DST Global closed in 2024 and a total war chest of one hundred seventy million, the company can execute on all four fronts simultaneously.

Two Analytics Engineer roles in London signal that the data foundation for Swap Inventory (the product a thirty-seven-million-euro UK raise funded) is still being hardened. Swap Inventory aims to deliver AI-driven restocking recommendations, pricing modeling, and demand forecasting that feed directly into the agentic storefront. An analytics hire today lets the feedback loop between real-time sales signals and inventory decisions tighten from batch overnight jobs to something closer to intraday. That loop is the difference between a storefront that reacts and one that anticipates.

In New York, the Senior Demand Generation Manager role anchors the go-to-market side of the agentic storefront launch. The job description for the partner-focused product marketing lead states plainly: "Defining how Swap wins. Shaping the Agentic Commerce category. Driving measurable revenue impact across products." Category creation is a marketing motion, but it only works if the product delivers the differentiation the narrative claims. The demand gen hire suggests Swap expects the storefront to be in market and selling (not just demoed) within two quarters.

The Partner Manager role for agency and technology partnerships, based in Austin, shows the distribution strategy. Swap's platform already covers cross-border, tax, returns, and demand planning; the agentic storefront brings new surface area that agencies and system integrators must learn to implement. "Establish and align with key partners for ongoing learning and enablement opportunities to ensure Swap's product and roadmap drives value for shared clients", that line from the job posting is a roadmap commitment: the product will evolve based on partner feedback, and the partner team owns that feedback loop.

The onboarding lead in London and a pair of support roles in Austin flag a scaling bottleneck the company knows looms. More than eight hundred brands on the platform today; each new storefront deployment adds configuration complexity across tax jurisdictions, return rules, and carrier integrations. Onboarding and support hires are the operational shock absorbers that let engineering stay focused on the agentic layer instead of drowning in implementation tickets.

The IT admin in Austin is the quiet tell: internal headcount grows fast enough that corporate infrastructure demands dedicated ownership. A 250-person company spread across five time zones outgrows ad-hoc IT.

Swap's careers page puts it bluntly: "We're building the agentic commerce OS, unifying logistics, returns, payments, and tax on a global scale." The hiring plan turns that unification concrete. Analytics engineers supply the intelligence. Demand gen and partner managers supply the channel. Onboarding and support teams supply the operational capacity to deliver at scale. The storefront (announced publicly as a brand agent that lets customers discover and purchase through conversation) forms the user-facing tip of that stack.

If hiring holds to plan, the next twelve months will bring Swap Inventory to general availability, the agentic storefront to a priced SKU, and the partner program to co-selling. The risk is integration: each layer has a different technical lineage, and the "unified OS" claim holds only if the seams don't show to the merchant. The next wave of hires (likely solutions architects and forward-deployed engineers, though none are posted yet) decides whether Swap ships a platform or a portfolio.

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