Josep Mingot

The through-line

Everything Josep Mingot says circles back to one idea: the most expensive problem in American healthcare sits in the administrative layer, not the clinical one. Because the payer and the caregiver are separate parties, care is preceded and followed by a chain of phone calls between practices, patients and insurance companies, and that chain is where delay, cost and confusion accumulate [1]. Prosper's answer is voice agents that automate those calls, both the ones patients place to a practice and the back-office ones the practice places to insurers [1]. He frames the ambition beyond automation: not just building patient assistants on the surface, but going after "these deeper workflows" that "actually reduce the friction that many patients have today in the healthcare, especially in the US, uh with insurance companies" [1]. That vision has been described publicly as making a trip to the doctor as simple as ordering a pizza, and the business behind it now automates thousands of medical calls [2].

The second through-line is about what actually makes such a company defensible. His view is that the product is not software people configure themselves, it is an outcome delivered through a deployment motion and a review loop, and that this is far harder to copy than a platform [1]. The company announced a $30 million Series A led by Andreessen Horowitz, and the conversation around it covered growth of 40 to 50% month on month and the playbook that landed the round [1][4].

On the friction between payer and caregiver

Mingot describes US healthcare organizations as running two parallel machines: teams facing patients, and back-office teams handling the interactions with insurance companies [1]. The problem is not only that the second machine is expensive, it is that it is invisible to the patient. Care requires "multiple touchpoints um with the insurance company that not just creates friction, it creates frustration in patients because many times it's not understood exactly what's going on" [1]. Patients do not know why care is delayed, why they are asked so many things before being seen, or why they end up owing a significant amount "without transparency or visibility before of of uh what was going on" [1]. Internally the team calls these delayed, siloed interactions friction along the patient journey, and the stated test of success is whether all of it can be streamlined end to end [1].

On staying on the administrative and financial side

He is emphatic about the boundary. Prosper is "super focused on the administrative side, on the financial side" and does not enter diagnosis or treatment [1]. The reasoning is partly about scope discipline: the administrative process alone, from a patient calling to book, through benefit checks and coverage verification, through charging insurer and patient, is complex enough on its own, and it spans the phone plus APIs, insurer portals and websites [1]. But the boundary is not a limit on ambition. His argument is that a patient-side assistant is only useful if it is also insurance savvy, capable of handling the interactions that today sit with separate teams, because otherwise the friction simply moves rather than disappearing [1].

On building a workforce of agents

The internal language is a workforce of agents, each with a job [1]. Some are patient-facing, such as scheduling agents that handle appointment requests and billing agents responsible for explaining what is owed for care; others are insurance interaction agents that call payers to verify a patient's benefits before care is delivered [1]. Voice sits at the core, but the work is agentic in nature because the same task may run across phone, API and portal [1]. Which agents get built is a deliberate choice rather than a technical accident: he says the team strategically picks core agents "that are not just cost cutting, but also revenue generating for the organizations that we work with" [1].

On the buyer and the ROI

The customers are large outpatient practices, for example a dermatology group with 20 locations and 100 doctors that centralises its call centre, verification and billing into one operations team, plus the EHR vendors that run those practices [1]. The buyer is a director of operations in permanent firefighting mode: 30 to 40% staff turnover in an entry-level role, national unemployment at 3 to 4% making good hires hard, and sharp seasonality with Mondays far worse than the rest of the week [1]. Around 80% of appointments still start with a phone call, so an unanswered phone is lost revenue outright [1]. The cost case is real but gradual, with a 20-person call centre eventually needing perhaps half that as hiring stops and people are redeployed to the complex cases they never had time for; the more interesting half is revenue, from picking up consistently around the clock and letting patients cancel and reschedule over the weekend instead of turning into Monday no-shows [1]. Mingot ties this directly to the roadmap: demonstrating a lift in appointment rate and in provider utilisation, meaning how many of a physician's available slots get filled, "is directly pricing power, right? That we may have against horizontal solutions" [1].

On outcome-based pricing

Pricing has been evolving rather than fixed. Several customers have been moved onto an outcome basis, priced on appointments or on the number of calls or tasks managed depending on which agents are deployed, and the pattern is that the more product surface a customer takes, the more outcome-based the arrangement becomes [1]. In every case the pitch rests on a clear ROI against the existing cost structure, with the revenue lift stacked on top [1].

On partner-led distribution

Prosper sells directly to large practice groups and also through the EHRs and other software and service providers those practices already run on [1]. Mingot calls the second channel "a very important hack" because it collapses sales cycles: reaching a health system that would have taken nine months to close becomes far faster when an existing distribution relationship is leveraged [1]. The trade-offs are explicit. Partners take a discount so they can hold their own margin and their own pricing power, often ending up as the default all-in-one solution for their customers, and the vendor gives up margin and control in exchange [1]. His condition for accepting that trade is volume: because the margin is thinner, the company selects mostly enterprise vendors where the reduced margin buys much greater scale [1]. Launching a partner is also time-consuming and can be a distraction, so choosing which partners to vet and prove out matters, particularly at seed stage [1]. The strategy is deliberately not to bet on a single channel, on the view that direct and partner-led sales reinforce each other [1].

On why partners will not simply build this themselves

Asked whether cheap software will let partners cut Prosper out, the position is that the partners are more exposed than Prosper is, because software is easier than agents [1]. The distinction drawn is between selling a platform on which someone else builds agents and selling a result: "We're selling them a KPI and everything that needs to happen, especially the feedback loop" [1]. That loop means executing the calls and interactions, reviewing all of them, and continuously optimising until they work correctly for that specific organisation [1]. The conclusion is that the moat is "software plus our deployment motion", a combination that has proven hard to replicate [1].

Takeaways

• The wedge is administrative and financial workflows only, deliberately excluding diagnosis and treatment, because the admin chain from scheduling to benefit verification to billing is complex enough on its own [1].
• A patient assistant is not enough on its own; it has to be insurance savvy and absorb the back-office interactions with payers, or the friction just moves elsewhere [1].
• Prosper structures its product as a workforce of specialised agents, scheduling, billing and insurance verification, rather than one general-purpose assistant [1].
• Choose agents that generate revenue, not only cut cost: lifting appointment rates and provider utilisation is what creates pricing power against horizontal competitors [1].
• Selling through EHRs and other incumbent vendors turns a nine-month enterprise sales cycle into a fast one, at the price of margin and control, so only enterprise partners with real scale are worth the discount [1].
• Defensibility comes from the deployment motion and the review-and-optimise feedback loop around the calls, not from the platform, which is why partners are more exposed than Prosper is [1].
• The buyer's pain is operational fragility: 30 to 40% turnover, a tight labour market and Monday spikes, against a market where roughly 80% of appointments are still booked by phone [1].