Who Gets Hired and Onto Which Teams
Eloquent AI, a seed-stage AI operator company that grew revenue elevenfold in its first year, according to Eloquent AI's careers page, has posted twelve openings: nine in engineering, two in deployment, one in product, with experience requirements ranging from one to five-plus years. The company, which calls itself the operating system for businesses in the AI age, raised a $10 million seed round, Eloquent AI's careers page reports, signed major banks and fintechs, and now runs AI operators daily inside those customers' environments. Its hiring profile is narrow by design.
The engineering roles cluster at mid and senior levels: AI Engineer for AIOps & Infrastructure (5+ years), Senior Software Engineer for full-stack work, AI Engineer for multimodal LLMs (3+ years), AI Engineer for agents (3+ years), AI Engineer for platform (1+ year), and Forward Deployed Engineer. A seventh engineering listing appeared within the last three days. React, Python, AWS, and GCP appear most often across the postings. Deployment roles focus on getting AI operators live inside customer environments, primarily banks and fintechs. The single product role suggests a lean product organization, with engineering carrying significant product ownership. The Eloquent AI Fellowship Program provides an entry point, though its experience level is not specified publicly.
That senior-weighted bias aligns with broader research on effective AI team structures. expertshub.ai's data shows teams function best with roughly two AI/ML engineers for every data scientist, a 60:40 to 70:30 senior-to-junior split, and a 70:30 generalist-to-specialist mix. Eloquent's public roles reflect the engineer-heavy, senior-weighted end of that spectrum. The Forward Deployed Engineer role — paid $120k–$180k versus $150k–$250k for core AI engineering roles — sits at the customer-facing edge of that structure, translating platform capabilities into production deployments.
The AI Agent Engineer listing specifies proficiency with PyTorch and TensorFlow, experience fine-tuning and optimizing LLMs for inference, and familiarity with enterprise integrations across AWS, GCP, or Azure. Candidates must "prototype, experiment, and iterate quickly" while working "closely with customers to refine AI solutions." Geographically, the team operates from San Francisco and London with remote eligibility for U.S.-based engineers. The London presence includes two advertised roles: Agent Deployment Manager and AI Engineer, signaling an active hiring wave there.
The tech stack signals a production-first orientation. Python dominates the AI engineering roles; React appears in full-stack and forward-deployed contexts; cloud infrastructure spans both AWS and GCP. That dual-cloud footprint matters for enterprise customers in regulated sectors who often mandate specific providers. What emerges is a team built for a specific stage: past prototype, past product-market fit, now scaling the machinery that lets AI operators run reliably inside banks. The next section maps what that machinery pays.
What It Pays
Eloquent AI's compensation sits at the intersection of two forces: the AI talent premium that has pushed specialist pay 20–40% above general software engineering since 2020, and the cash constraints of a seed-stage company that raised $7.4M in September 2025 and graduated from Y Combinator that April. The board's live salary band runs $105k–$250k base with a median of $250k across eight salaried postings, a range that reflects both the market heat for ML talent and the company's current funding tier.
| Role | Location | Base Salary Band (USD/year) |
|---|---|---|
| AI Engineer, AIOps & Infrastructure | San Francisco | 150,000 – 250,000 |
| Senior Software Engineer, Full-Stack | San Francisco / Remote (US) | 150,000 – 250,000 |
| AI Engineer, Multimodal LLMs | San Francisco | 150,000 – 250,000 |
| AI Engineer, Agent | San Francisco | 150,000 – 250,000 |
| AI Engineer, Platform | San Francisco | 150,000 – 250,000 |
| Forward Deployed Engineer | San Francisco | 120,000 – 180,000 |
Source: Zero G Talent board postings (live)
Five of six roles cluster at $150k–$250k, Zero G Talent's board data shows. That clustering is deliberate. Industry data shows senior ML engineers at top labs now command total compensation exceeding $800k, and even entry-level AI roles in major hubs start above $100k. Eloquent AI's bands land squarely in the early-stage startup tier: competitive enough to attract engineers with production LLM experience (which carries a 15–25% premium over research-only backgrounds) but capped well below the $350k–$500k total packages that Big Tech offers for equivalent seniority. The Forward Deployed Engineer role sits lower at $120k–$180k, Zero G Talent's figures put, consistent with a customer-facing technical role that blends implementation and solutions engineering.
Third-party estimates vary. Alion.io models ranges from $68k–$174k up to $167k–$317k across roles. Swiftcruit.ai lists ranges from $70k–$100k up to $150k–$250k. Salary.com puts the company-wide average near $103k with a $91k–$117k interquartile range. The spread reflects different methodologies (some model total compensation, others base only) and the noise inherent in a company with fewer than ten posted roles. The board's first-party data is the cleanest signal: eight roles, one consistent band.
Equity terms have not been published. At a $7.4M–$10M seed valuation, industry practice for early engineers falls in the 0.1%–1.0% range, vesting over four years with a one-year cliff. The math is clear: at a $50M valuation, 0.25% is worth $125k on paper; at $1B, $2.5M; at zero, nothing. Private equity should be evaluated at a 30–50% discount to paper value because it is illiquid and carries real downside risk. Eloquent AI has not published its option pool size or refresh policy. Candidates should ask directly about the current valuation, outstanding shares, and whether annual refresh grants exist; companies like Google and Meta provide refreshers worth $50k–$200k yearly, but most seed-stage startups do not formalize them.
Benefits listed publicly include a company offsite in Italy. The careers page mentions "competitive salary and meaningful equity" but does not detail medical, dental, vision, 401(k), or learning budgets. The board lists one role as "San Francisco / Remote (US)"; the careers page shows five additional remote-eligible roles including a London position and a Europe-based front-end role. Candidates negotiating from lower-cost metros should know the industry trend is toward less aggressive location adjustment, but exceptions are made case by case.
Total compensation at this stage is base-heavy. Industry data puts base at 40–60% of total comp at mature companies; at Eloquent AI, with equity unpriced and no public bonus structure, base is effectively the guaranteed portion. Signing bonuses have not appeared in posted bands. The revenue signal matters: third-party sources report $500k ARR across five employees ($100k revenue per employee) alongside the elevenfold growth claim, implying the next funding round will reset the valuation. Anyone joining now prices their equity against that inflection. The board shows 12–13 open roles and stable hiring momentum (51/100). The company is spending its seed capital on talent. The offer you negotiate today will be benchmarked against the next round's terms — not the last one.
How the Hiring Process Works
The company has not published its interview funnel, and no first-party accounts from recent candidates appear in the public record. The company's careers page lists open roles but does not detail stages, timelines, or evaluation rubrics. General research on frontier-AI labs of this profile suggests a multi-stage technical evaluation, but no Eloquent AI-specific process is documented. Candidates should verify each step directly when engaging with the company.
Where the Work Happens — and Who Thrives There
Eloquent AI operates from a single headquarters in San Francisco, the address listed across its incorporation filings, accelerator record, and job postings. The company was officially filed on July 8, 2025, with its corporate leadership (CEO Tugce Bulut, CFO Aldo Lipani, and Burce Bulut Ozan) all registered at San Francisco addresses. Its Y Combinator X25 batch membership means the founding team spent the first half of 2025 working out of YC's campus in the city.
The physical footprint matches the headcount. Third-party directories peg the team between seven and eleven employees as of mid-2026, a size that typically translates to a modest office. Built In San Francisco lists the workspace as "OnSite Workspace: Employees work from physical offices" but adds "Typical time on-site: None," a phrasing that signals a hybrid default rather than a mandate. The job board data bears this out: five of the six salaried roles posted to the Zero G Talent board are tagged "San Francisco" only, while the Senior Software Engineer, Full-Stack role explicitly lists "San Francisco / Remote (US)" and a Software Engineer, Front-End, Europe role points to a distributed hiring lane. The Forward Deployed Engineer role is also San Francisco-only.
The Europe listing hints at the next physical node. A single front-end role tagged "Europe" suggests either a remote-first hire in a friendly time zone or the seed of a future hub. Given the company's focus on banking and insurance across retail, commercial, and investment segments (sectors with heavy European regulatory regimes), a local presence would help with data-residency requirements and on-premises deployments the platform already supports. The website emphasizes "private deployment of Eloquent AI for ultimate data control, security, and compliance, seamlessly integrating into your virtual private cloud or on-premises environment."
For now, the San Francisco HQ is the anchor. The YC alumni network, the concentration of fintech prospects, and the density of AI talent in the city make it the logical base. The hybrid posture (office available, on-site time optional) reflects a team small enough to coordinate directly but building a product complex enough to need whiteboards, GPU access, and the kind of spontaneous debugging that still happens faster in person.
The culture that fills that space is explicit. The careers page states it plainly: "People with agency." That phrase does the heavy lifting for an eleven-person company that achieved elevenfold revenue growth in year one, secured major bank and fintech clients, and now operates an AI Operator trusted to automate up to 96% of complex, regulated customer operations without engineering effort or API integrations. The traits that follow are not aspirational — they are survival requirements.
Ownership that spans the stack. The careers page lists the first criterion: "Take full ownership of what you build, from infrastructure to data integrity." At this size, there is no platform team to hand off to, no SRE rotation to absorb operational pain, no data engineering function to clean up schema decisions. The AI Engineer who ships a multimodal LLM feature also owns the inference infrastructure, the data pipeline feeding it, and the integrity guarantees that regulated customers (banks, insurance firms, debt-recovery operations) audit. The Forward Deployed Engineer who sits with a customer in London or San Francisco owns the integration into that customer's Intercom, Zendesk, Salesforce, or Stripe environment end to end. "Ship independently without needing supervision" is not a perk; it is the only way the math works.
Problem-first, code-second. "Understand the problem before writing code" reads like generic advice until you see the domain: financial services customer operations where a hallucinated refund, a misrouted compliance escalation, or a leaked PII field triggers regulatory action. The AI Operator "learns to securely navigate your existing systems simply by observing your standard operating procedures." That means the engineers building it must model SOPs, not just APIs. They must understand why a cash-advance workflow differs from a debt-recovery workflow, why NICE CXone integration matters for a contact center, why Alloy and Unit21 appear in the integration list alongside Gmail and Slack. The sectors span banking, insurance, fintech, cash advance, and debt recovery, each with distinct compliance surfaces. Candidates who reach for a framework before mapping the regulatory boundary will not last.
User and impact fluency. "Know the user, the objective, and the impact" maps directly to the product's measurable claims: up to 96% automation of customer issues, 36% uplift in inbound conversions, fourfold cost reduction. The user is not an abstract "customer" — it is a compliance officer at a major bank who needs audit trails, a support lead at a fintech who needs first-contact resolution, a debt-recovery manager who needs every interaction logged and defensible. The engineer who cannot translate a model latency improvement into a compliance-team time saving, or a multimodal capability into a reduced drop-off rate, is optimizing the wrong variable.
AI tooling with accountability. "Use AI tools intelligently. Review and stand behind every line they produce." This is a company building AI Operators; its engineers use coding assistants, LLM-based test generation, and automated refactoring daily. But the same regulated environment that demands 96% automation accuracy demands that the engineers themselves never ship code they cannot explain or defend. The culture selects for practitioners who treat AI as a lever, not a crutch, who can prompt, verify, and revert with equal speed.
Subtraction over addition. "Remove complexity. The best systems are built by subtracting." The product pitch emphasizes "no engineering, no APIs required" for customers. Internally, that philosophy means the team builds primitives that eliminate whole classes of integration work rather than accumulating adapters. A candidate who measures productivity by lines of code or service count will fight the culture. The ones who thrive measure progress by dependencies removed, SOPs codified, and customer engineering hours saved.
High-autonomy, high-growth tolerance. The "Why join" section lists "Full ownership. Small team. High autonomy. You own what you ship end to end. Hyper growth. 11x ARR in a year. Demand is accelerating fast." Twelve open roles across San Francisco and London mean the team could double. The people who join now will define the architecture, the hiring bar, and the cultural norms for the next fifty hires. They will also absorb the chaos of an early-stage company that already counts major banks as customers. That combination — regulated-customer rigor at pre-Series A headcount — filters for a specific profile: engineers who have operated in constrained environments (fintech, health tech, defense) but have not calcified into process-heavy organizations.
Regulated-industry fluency. The integration stack (Okta, Alloy, Unit21, Prelim, NICE CXone) signals the compliance surface. Candidates who have implemented SOC 2 controls, worked with PCI-DSS scope, or built audit trails for financial regulators enter with a vocabulary the team otherwise has to teach. Eloquent AI is trusted by leading enterprises and regulated companies for providing the most reliable and secure autonomous AI Operators that communicate with accuracy and integrity. That trust is earned per deployment; the engineers who sustain it treat security and compliance as product features, not checkbox exercises.
Distributed, in-person collaboration. Two offices (San Francisco and London) with an offsite in Italy. The team is small enough that time-zone overlap matters; the Forward Deployed Engineer role remains explicitly San Francisco-based, while the Senior Full-Stack role lists "San Francisco / Remote (US)." The culture selects for people who can build rapport face-to-face during the offsite and sustain it asynchronously across eight hours. Candidates who need daily stand-ups to stay aligned will struggle; candidates who write clear RFCs, record concise Looms, and default to over-communicating decisions in writing will thrive.
The through-line is agency in a constrained, high-stakes environment. Eloquent AI does not hire for potential in the abstract — it hires for demonstrated ownership in domains where mistakes are expensive and autonomy is non-negotiable. The pyramid the company inverted on day one (senior-heavy, engineering-led, no interns) is now the filter for the next fifty hires. Whether that structure holds at Series A scale is the only variable the current team cannot control.
Working in AI? Zero G Talent tracks the openings: see every open Eloquent AI role, browse AI jobs, the companies hiring, and the people building the field.