How Work Gets Done
A two-year-old company has landed a $25 million Series A led by Bessemer Venture Partners, secured strategic checks from Vanguard, Guardian Life, and SEI, and deployed its platform across more than 1,100 wealth professionals at Mercer Advisors powering hundreds of thousands of discrete actions annually. That compression — idea to enterprise deployment in roughly two years — sets the cadence for everyone inside the building, and the buildings, plural.
The team sits at roughly 36 to 38 people (Levels.fyi counts 36; Built In shows 38; LinkedIn's 51–200 band is stale). They operate from two primary hubs, New York and Bern, Switzerland, with a production support specialist in Novi Sad, Serbia, and a senior DevOps engineer hired remote. That geographic spread isn't incidental; it's structural. Co-founders Bassam Chaptini and Rabih Ramadi serve as co-CEOs, distributing final authority between two technical founders rather than concentrating it in one. Ramadi has described the platform's AI agents as proactive collaborators: "The AI agent can be like, 'Hey, you haven't talked to this client in eight months, we should trigger this specific action. And by the way, based on public data sources and email interactions, that client just had a child. You should suggest they open a 529 plan.'" Chaptini frames the core problem as fragmentation: "It's all extremely fragmented. It's a headache for people to keep on marrying that data." Their shared conviction, that AI adoption works best when built around a firm's operating model, not layered on top, doubles as an internal operating principle.
Decision-making follows the same logic. With no middle-management layer visible in the org chart, product and engineering calls route directly to the founders or to the senior hires they've brought in recently: Hemang Bhojani as Head of Wealth Solutions and Thomas Moore as Head of Wealth Go-to-Market, both announced in early 2026. The job board tells the rest of the story. Open roles (Principal Product & Client Solutions, Principal Designer, Senior Product Designer, Account Director Wealth) carry "Principal" and "Senior" titles at a 36-person company. That isn't title inflation; it signals that individual contributors own outcomes end-to-end, not tickets. The Principal Product role sits in New York. The Principal Designer sits in New York. The Senior DevOps Engineer works remote. The production specialist sits in Novi Sad. Authority follows expertise, not geography.
The pace shows in the partner integrations. Guardian's partnership, announced February 2026, promises "next-generation technology to deepen client relationships" with AI tools rolling out to advisors across investment and protection needs. Vanguard and SEI participated in the Series A not as passive limited partners but as strategic backers signaling they expect the platform to embed inside their advisor workflows. Mercer Advisors, already live, describes Avantos as having "transformed how our advisors meet their needs" by unifying "multiple disconnected systems into a single environment." Each integration demands security reviews, compliance alignment (the platform is SOC2 certified), and data-mapping across custodians, CRMs, and core systems. At Avantos, engineers who build the knowledge graph also own custodian integrations and the advisor-facing workspace.
That concentration of scope per person creates the culture's defining tension. Autonomy is real because it has to be: there's no one to delegate to. The co-CEO model means strategic disagreements get resolved in a room of two, not escalated up a chain. The distributed footprint makes synchronous meetings expensive, so context moves through writing (specs, RFCs, async updates) not stand-ups. The senior titles on the job board mean new hires join as peers to the early team, not subordinates. For a candidate, the question isn't whether the culture is "good" or "bad." It's whether you can operate without a net — because at 36 people moving at this velocity, there isn't one.
Operating Principles
Avantos.ai does not publish a values page. Its operating principles are legible in the decisions Chaptini and Ramadi have made since they began building in stealth with Mercer Advisors — decisions that consistently favor technical honesty over marketability, customer proximity over distribution, and architectural durability over speed to label.
The founding sequence itself is a principle stated in action. Chaptini and Ramadi spent 20 years across two companies watching financial services firms stitch together CRMs, task managers, portfolio systems, and paper. They did not incorporate and then hunt for a beachhead. They identified the white space, secured Mercer Advisors as a design partner, built the product in stealth, deployed it, measured a 30 percent lift in advisor-to-client ratio, and only then founded the company in September 2024. The principle: validate with a live, urgent buyer before the entity exists. Chaptini put it directly on the Frontlines podcast: "Once we started getting the impact, the real impact by deploying it, is when Rabih and I decided to found the company."
That same discipline shaped the architecture. The team started with relational tables, the default choice for structured data. The model collapsed under the weight of modeling clients, service teams, and complex products simultaneously. "It gets out of control very quickly. It doesn't matter how you structure it. It does so very quickly. Hence why people revert to paper a lot of the times." The move to a knowledge graph was not a technology bet; it was a constraint surrender. The problem demanded a structure that could represent multi-entity relationships without schema explosion. The graph delivered that, and it turned out to be the native substrate for the AI agents Avantos layered on top, agents that require the rich contextual data a graph produces. The principle: let the data model follow the domain complexity, then build the AI on the substrate that actually supports it.
The go-to-market motion carries the same fingerprint. Avantos reached Mercer, Guardian Life, Vanguard, and SEI without a marketing team, without inbound, without demand generation. Chaptini's explanation is blunt: "Going to enterprise and asking to be a foundational layer does not require marketing. It requires you to know the client super well. It requires a network for you to come in and be credible, obviously with the track record you have." The round was deployed to recruit design partners in each adjacent vertical: Guardian Life for insurance, Vanguard for brokerage-to-wealth cross-sell, SEI for banking, each already operating a wealth business that Avantos could expand from. "We're picking the design partners that are very strategic and using wealth to expand into the other products." The principle: enterprise foundational sales run on credibility and co-building, not funnel mechanics.
Category positioning follows the same logic. "CRM" was captured by Salesforce and redefined as a sales tool. Chaptini refused the framing war: "We show up and we say we focus on client management, particularly onboarding and servicing. Our clients immediately get it." When the pain is acute, the category can wait. The principle: lead with the workflow the buyer already knows is broken.
Security and compliance are not afterthoughts. It is SOC2 certified and "architected for enterprise security, compliance, and auditability from the ground up." Integration is explicit: "Integrate with your CRMs, custodians, and core systems so you can adopt Avantos without replacing what you depend on." Customization is promised without engineering overhead: "Configure onboarding and servicing journeys to match how your firm works." The principle: meet the enterprise where it lives — secure, integrated, configurable — rather than demanding a rip-and-replace.
The automation narrative is deliberately reframed. Enterprise buyers arrive with two fears: data isolation and "what happens to the people." Chaptini answers the second by reframing the constraint: "The binding constraint isn't headcount — it's advisor capacity." Avantos measures success by advisor-to-client ratio, not headcount reduction. A 30 percent increase means more clients served, more AUM, more revenue. The principle: expand the scarce resource (advisor capacity) rather than automate the abundant one (tasks).
These principles compound into a culture the job board signals reflect. Open roles span Principal Product & Client Solutions and Principal Designer in New York, Account Director Wealth in New York, Production Support Specialist there, Senior DevOps Engineer remote, Senior Product Designer in New York; a distributed footprint that demands the autonomy the founding decisions encoded. The operating system is not written on a wall. It is written in the sequence: design partner before incorporation, knowledge graph before AI layer, credibility before marketing, pain before category, capacity expansion before cost cutting. Candidates who need a values deck to know how to act will wait for one that never arrives. The people who stay are the ones who read the decisions and start operating.
The Hiring Bar
Avantos.ai's hiring filter reads like a proxy for the company's own constraints: a distributed team building a knowledge-graph platform that sits inside the operating model of wealth-management firms, not on top of them. The research signals converge on a single profile, someone who has already operated at the intersection of financial-services workflows and modern AI systems, and who can move without a detailed spec.
The company states explicitly that "scaling AI in wealth management and broader financial institutions takes people who've done the work from the inside." That sentence, posted to LinkedIn in September 2026, functions as a de facto hiring manifesto. It rules out pure technologists who have never seen a custodial feed or a compliance review queue, and it rules out industry veterans who treat AI as a vendor feature request. The roles currently open (Principal Product & Client Solutions, Account Director Wealth, Senior DevOps Engineer, Senior Product Designer, Principal Designer, Production Support Specialist) all sit at that seam. The Principal Product role, listed at an hourly rate suggesting a specialized contract engagement, calls for someone who can translate advisor pain into product logic. The Account Director role demands credibility with RIAs and enterprise wealth firms. The DevOps and Designer roles require shipping in a remote-first, high-cadence environment where the platform already powers that volume annually across that many professionals at Mercer Advisors alone.
The "Why Join" page on the company's careers site reinforces the same filters: "Massive influence. Your insights directly shape client success and the product roadmap. Strategic ownership. Lead core enterprise clients end-to-end." Those are not perks; they are selection criteria. A candidate who needs a product manager to write tickets will not survive the interview loop. A candidate who has never owned a client relationship from onboarding through tax-prep workflows will not pass the domain screen. The company's own language, which emphasizes building around that same principle rather than layering on top, doubles as a technical litmus test. Engineers and designers who default to generic SaaS patterns fail it; those who have wrestled with configuration-based application models and user-defined journeys pass.
The senior hires announced in September 2026 — Hemang Bhojani and Thomas Moore, leading Wealth Solutions and Wealth Go-to-Market respectively — illustrate the bar in practice. Both bring operating experience from inside wealth-management firms, not just adjacent tech. The company's stated future, scaling AI agents across the advisor stack, expanding across RIAs, independent broker-dealers, and enterprise wealth firms, deepening client relationships without adding operational complexity, requires people who already know where the complexity lives. The hiring process does not publish its rubric, but the evidence is consistent: it selects for operators who have felt the friction of fragmented systems, who can write code or design flows that reduce that friction, and who will drive a workstream end-to-end in a distributed team with no one checking their calendar. Candidates weighing the offer are really weighing whether they can carry that load without a safety net.
What Employees Say
The public record on Avantos.ai is thin and complicated by name confusion. Glassdoor lists reviews for "Avantos," "Avantus," and "Avantio," three distinct spellings that may represent the same entity at different points or entirely different companies. The distinction matters for anyone weighing an offer.
The earliest detailed review, dated November 2018, appears under "Avantus" (E419934). That reviewer gave a 1.0 rating and wrote: "Literally a dumpster fire." They added that "a few good people there are trying their best to make progress" and described being hired for a data science role "due to the CEO's obsession with AI/Machine Learning/the Singularity/anything with a buzz word attached to it." The review suggests a chaotic, hype-driven environment six years ago. Whether that entity became Avantos.ai or simply shares a similar name is not documented in the available sources.
A separate Glassdoor page for "Avantio" (E1790617) shows a 4.2 out of 5 rating across 20 reviews, described as indicating "most employees have an excellent working experience there." The similarity in spelling has likely inflated or confused aggregate scores on some aggregators.
The only review explicitly tied to "Avantos" (E5182083) is dated January 15, 2025. It carries a 4.0 rating and comes from a current employee with more than three years' tenure. That person wrote: "I am super flexible with my time." The only downside noted: "Sometimes calls with co-workers take longer than expected but thats about it." The review is brief, positive, and consistent with the high-autonomy, distributed model described elsewhere in this article, but it is a single data point.
Glassdoor's overview page for Avantos shows one review total as of the 2025 date. The "See All Reviews (34)" link on the Avantus page suggests a larger historical corpus, but those reviews sit under the Avantus spelling and predate the 2025 Avantos review by six years.
First-party board data confirms Avantos.ai is actively hiring: recent postings mirror the roles listed above. The geographic spread (New York, Novi Sad, remote) aligns with the distributed structure the 2025 reviewer alluded to.
The picture that emerges is fragmented. A negative 2018 account under a similar name. A strong aggregate score for a different company (Avantio). One positive, recent review from a long-tenured employee at the actual Avantos.ai. Candidates should treat the 2018 feedback as historical context for a possibly related entity, discount the Avantio score entirely, and weigh the single 2025 review as the only direct, current signal, while recognizing that one review, however detailed, cannot capture the full range of experience inside a growing team.
Who Thrives, Who Burns Out
The profile that succeeds at Avantos is shaped less by resume keywords than by the structural realities of a 38-person, Series A company building an AI-native knowledge graph for wealth-management enterprises. The team is distributed across New York, Novi Sad, and fully remote nodes, which means the default mode is asynchronous, written, and low on ceremony. Someone who needs a daily stand-up to know what to do next will stall; someone who writes a crisp spec, ships a prototype, and iterates from customer feedback without asking permission will accelerate. The product itself — unifying onboarding, servicing, and relationship intelligence into a single context layer for firms like Mercer Advisors, Guardian, Vanguard, and SEI — demands both depth in distributed systems and comfort with regulated-enterprise constraints. SOC2 compliance, custodian integrations, and household-entity data models aren't optional learning curves; they're the daily substrate.
Candidates who thrive tend to have operated in small, high-leverage teams where the distance between idea and production is measured in hours, not sprints. They've felt the pressure of a Bessemer-backed Series A timeline: the board expects velocity, and the enterprise pilots expect reliability; they treat that tension as energy rather than anxiety. Those role listings reveal the same pattern: ownership of infrastructure and reliability without a dedicated platform team to fall back on. Design roles (Senior Product Designer and Principal Designer, both New York-based) imply a product culture where design isn't a handoff step but a co-leader of discovery, likely working directly with the founding engineers and early customers.
The burnout profile is the mirror image. People who need frequent synchronous alignment, who expect a manager to decompose tickets, or who treat compliance and security as "someone else's job" will exhaust themselves trying to keep up. The distributed footprint means no hallway conversations to clarify ambiguity; you write the RFC, you gather async feedback, you decide. The enterprise client list means no "move fast and break things"; a broken onboarding flow for Mercer Advisors is a reputational event, not a bug ticket. And the 38-person headcount means every role is a force multiplier: there is no backend team, no frontend team, no DevOps team; there are engineers who own outcomes end to end. Candidates who have only worked in specialized functions at larger companies often underestimate the cognitive load of context-switching across the full stack, the customer, and the regulatory frame simultaneously.
The hiring bar, described in the previous section, selects for this profile implicitly. But the filter is imperfect: some candidates pass the technical screens and values conversations only to discover, three months in, that the autonomy they celebrated in the interview feels like abandonment in the daily grind. Those who stay are those who build their own scaffolding (personal OKRs, async communication rituals, direct relationships with the enterprise champions at Guardian or Vanguard) and treat the lack of guardrails as the job itself. The net isn't missing. It's just yours to weave.
Working in AI? Zero G Talent tracks the openings: see every open Avantos.ai role, browse AI jobs, the companies hiring, and the people building the field.