The Problem Ember Solves
Ember AI lists four AI-focused roles in Burlingame, and the specialists applying are learning the hard way: the screen prioritizes one thing — proven success automating healthcare revenue-cycle tasks.
Ember AI, a Y Combinator-backed revenue-cycle platform founded in 2022 (F24 batch), runs a 25-person team that has grown 200 percent month-over-month. CEO Charlene Wang, former Google Healthcare AI Product Lead, and Warren Wang, MIT AI researcher with NVIDIA and CSAIL experience, have said the problem compounds as specialty groups expand across providers, locations, and payer contracts: coding consistency and revenue protection grow exponentially harder. Eye care — particularly retina — sits at the sharp end, where high-cost injection management and payer-specific rules create administrative drag that directly threatens margin.
The platform layers onto existing EHR and practice-management systems (Nextech, Veradigm, athenahealth, ModMed, eClinicalWorks, NextGen, Epic, Oracle Cerner, and others) without custom development. Customers submit a few hundred cases and receive initial analysis in three days. Three engines share a single data foundation. The Foundation Data Engine unifies clinical documentation, coding decisions, payer policy, and contracts across every encounter. The Audit Coding Engine reviews every encounter against national standards, payer-specific policies, internal guidelines, and contracts; every flag carries a rule citation. The Recovery Appeal Engine reads CARC and RARC codes, identifies applicable LCD and NCD policy and contract terms, drafts appeal letters, packages documentation, and tracks claims through adjudication.
This architecture powers what Ember calls "agents for every step of the revenue cycle" on an AI orchestration layer. Pre-bill audit catches undercoded and overcoded claims before submission. Denial intelligence identifies patterns and root causes. Appeals automation pulls records, references policy, drafts the packet, and pushes it through tracking. Contract intelligence benchmarks payer rates to surface underpayments. The company says the result cuts manual rework by up to 80% and reduces denial rates by more than half.
Early data from Midwest Vision Partners, a multi-state eye-care network, backs those claims. Announced as a strategic partnership on August 18, 2026, the engagement ran Ember across more than 200,000 encounters in its first three months. MVP uncovered over $1 million in missed revenue, PRNewswire reported. The engine surfaced more undercoding than overcoding, recovering revenue that would have gone uncaptured while protecting against compliance exposure. MVP's VP of Physician Engagement, Dr. Katie Greiner, said Ember delivered what it promised and then some, noting responsiveness as a differentiator: when MVP raised an issue, the team brought solutions rather than waiting to be asked. Measured outcomes Ember reports: a 23 percent improvement in clean-claim rate within 30 days, 57 percent fewer denials, 3.1 times faster appeal turnaround, and every denial worked.
Four Roles, One Screening Bar
Ember's careers page lists four open positions as of July 2026, all based in Burlingame, California, with a mix of on-site and hybrid arrangements. The roles span engineering, growth, and sales, reflecting the team's growth since its Y Combinator F24 batch and now works with physician groups and health systems across the U.S. to increase net revenue by 16 percent through AI-driven revenue-cycle automation.
Software Engineer, Full Stack (On-site only). This is the lone pure engineering role on the public careers board, distinct from the "Founding Engineer" listing on Y Combinator that carries a $140,000–$180,000 base plus equity and had attracted over 200 applicants by late July. The full-stack position sits at the center of Ember's technical challenge: building agents that navigate the fragmented, decades-old software stacks of payers and providers. The founding team has signaled that eval infrastructure for healthcare-specific workflows is a core problem area.
Growth Marketer (Hybrid). This role owns the top-of-funnel engine for a product selling into a buyer universe where most doctors "don't have that much time" to evaluate contracts they see once or twice a year, and where providers often believe they're "too small to be important" for value-based arrangements. The marketer must translate Ember's 16 percent revenue-lift claim into language that cuts through payer-provider noise, targeting physician groups and health systems drowning in administrative complexity but skeptical of yet another AI pitch. Experience with B2B healthcare SaaS cycles, especially RCM or value-based care, is the unstated filter.
Account Executive, Mid-Market (Hybrid). This carry-quota role sells into mid-market physician groups and health systems — organizations large enough to feel RCM pain acutely but too small to build internal AI teams. The sales motion is consultative: Ember's own research notes that "signing the contract is actually the easy part; the hard part is then going to every provider on a daily basis if not weekly basis educating them on what the requirements are." The AE must navigate multi-stakeholder deals where clinical staff, billing teams, and compliance officers all veto differently. Prior experience closing six-figure deals into provider organizations, with a track record in RCM or clinical documentation improvement, is the baseline.
Business Development Representative (On-site only). The BDR role feeds the AE pipeline in a market where outbound requires fluency in CMS rule changes, including risk adjustment moving from v24 to v28, Stars and HEIDAS shifting to outcomes-based, special-needs-plan complexity, and the audit and compliance pressure trickling down from payers to providers. The on-site requirement reflects a team culture that emphasizes "in-person culture: build alongside the team at our Burlingame office" and daily lunch as a collaboration mechanism.
Across all four roles, Ember's stated values — "Do what we say," "Own the outcome," "Raise the standard" — function as the real screening rubric. The company's public posts emphasize that "everyone is a joy to work with" and point to a referral program with a dedicated Google Form, suggesting internal networks are a primary sourcing channel. Compensation details beyond the founding-engineer band are not public, but the benefits package lists comprehensive health coverage, professional-development budget, office-setup stipend, daily meals, 10 paid holidays, and an annual offsite.
What Applicants Are Doing
Ember's Y Combinator job posting states the company is "growing very fast," a pace that typically correlates with a sharp uptick in inbound applications. The public signal is visible on LinkedIn, where a search for "Ember AI" returns 15 current U.S. listings while a broader "AI Ember" query surfaces 166, suggesting the company has been posting aggressively across multiple titles and locations over recent weeks. Neither figure breaks out application counts, and Ember has not published funnel metrics for the four roles highlighted in this hiring sprint.
A cluster of AI-powered resume tailoring services (Jobifer, Resumaier, and TailorResume) now market themselves explicitly around the problem of matching candidate experience to specific job descriptions in seconds. Their pitch is that generic large language models miss the nuance of specialized domains, a claim Ember's own hiring blog validates from the staffing side: "Generic AI tools don't know what 'CVICU RN' means. They can't distinguish between a candidate who was genuinely unavailable six months ago versus one who simply didn't respond."
The absence of disclosed application numbers, conversion rates, or time-to-fill metrics means any claim of a "surge" rests on inference rather than measurement. What is documented is the emergence of a tooling layer built to help candidates perform keyword and narrative alignment, and Ember's own warning that generic versions of those tools fall short on healthcare semantics.
Why This Hiring Pattern Matters
The screening criteria (documented revenue-cycle impact, hands-on automation of claims workflows, fluency with payer-specific denial patterns) mirror a shift that has moved from pilot programs to board-level mandate across the provider landscape. The HFMA's May 2025 poll found 63 percent of healthcare organizations already using AI and automation in the revenue cycle, with 48 percent applying it to documentation and coding, the leading use case. Seventy-three percent expect the biggest impact on prior authorizations; 67 percent point to denials and underpayment management. Payers, by most accounts, are ahead. DHInsights noted in January 2024 that providers are playing catch-up, and the mid-revenue cycle is where the financial pressure (rising labor costs, margins below pre-pandemic levels, denials on an upward swing) makes AI adoption existential rather than experimental.
The talent market has not kept pace. Deloitte's 2026 State of AI in the Enterprise survey identifies insufficient worker skills as the single largest barrier to integrating AI into existing workflows. Education, not role redesign or hiring, was the number-one adjustment companies reported making to their talent strategies. Yet the same research shows that the most successful organizations do not stop at upskilling; they reimagine jobs to combine human judgment with AI capability, creating roles such as AI operations managers, human-AI interaction specialists, and quality stewards. Organizational charts are flattening as routine execution shifts to models, and some systems are merging technology and people-leadership functions so that workforce design evolves alongside the stack.
Revenue-cycle leaders describe the same dynamic in more concrete terms. Atlas Healthcare Partners cut labor dependence 31 percent and denials 48 percent across 28 ambulatory surgery centers by targeting the front-end processes (registration, eligibility, prior authorization) that generate the majority of back-end denials. The goal, repeated across surveys, is zero-touch claims: the most efficient way to get paid is without human intervention at all. But reaching that rate requires people who understand both the clinical coding logic and the automation layer that sits on top of it. Akasa's 2023 trend report noted the industry is moving from computer-assisted coding toward autonomous coding, while acknowledging fully autonomous remains out of reach. Coders are being redirected to auditing and complex chart review, work that demands deeper domain expertise, not less.
The hiring implications are structural. Nearly 60 percent of hospitals and health systems carry 100 or more open roles across operations. Deloitte estimates current technology can free up to half of revenue-cycle professionals' time and one-fifth of bedside nurses' time, translating to hundreds of hours per employee per year. But capturing that capacity depends on hiring for a hybrid profile: part revenue-cycle operator, part AI-literate workflow designer. The HFMA poll lists IT infrastructure limits (51 percent), budget (44 percent), integration challenges (43 percent), and ROI demonstration (42 percent) as top obstacles, all problems that sit at the intersection of technical and operational knowledge.
Governance is the next bottleneck. Only one in five companies has a mature model for overseeing autonomous AI agents, according to Deloitte, even as agentic AI usage is poised to rise sharply. HFMA sources stress that AI governance will be equally important as the technology scales, and organizations must be prepared to answer regulatory and standards questions as they evolve. That creates demand for a third competency: compliance-aware AI deployment within healthcare's specific regulatory envelope.
As MVP's Dr. Katie Greiner put it, Ember "did so" — the kind of outcome the market now hires for. The four roles in Burlingame remain open. The screen waits for the resume that proves it has already automated a denial appeal, cut a prior-auth turnaround, or recovered revenue a human missed. The rest is noise.
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.