Mandolin vs Thoughtful AI pricing: how do they typically charge (per workflow, per volume, per site) and what’s the ROI vs adding FTEs?

Most specialty-drug leaders don’t actually care whether a vendor charges “per workflow” or “per bot.” They care if the math beats adding another FTE, another outsourcing contract, or another year of backlog and denials. The real comparison between Mandolin and Thoughtful AI is how each model maps to your volumes, sites, and staffing—and how predictable the ROI is when you’re living in payer portals, faxes, and phone calls.

Below, I’ll break down how Mandolin typically structures pricing versus a Thoughtful-style RPA model, what “per workflow / per volume / per site” really looks like in practice, and how the ROI compares to hiring or outsourcing more back-office staff.

I’m writing from the operator seat: I’ve run benefits verification, prior auth, and appeals teams for multi-site ambulatory infusion. I’ve done the FTE vs. vendor math in detail—down to minutes-per-referral, denials per 100 claims, and margin swings from 340B and GPO pricing.

---

Quick Answer: The best overall choice for high-volume specialty-drug back-office work is Mandolin. If your priority is general-purpose RPA across many non-clinical departments, Thoughtful AI is often a stronger fit. For low-volume, narrow workflows where you want to experiment with automation before committing, consider Thoughtful AI in a limited scope.

At-a-Glance Comparison

Rank	Option	Best For	Primary Strength	Watch Out For
1	Mandolin	High-volume specialty-drug operations (infusion, buy-and-bill, specialty pharmacy)	End-to-end specialty-drug workflows priced to replace large chunks of back-office FTE hours	Requires enough volume and specialty-drug complexity to fully realize ROI
2	Thoughtful AI	Broad RPA needs in finance, HR, or generic rev cycle	Flexible, bot-based automation that can touch many different back-office functions	Workflow setup and maintenance burden, especially in payer portals and fax-heavy workflows
3	“Add FTEs” (status quo)	Small, stable programs with low growth & simple access workflows	Direct control, familiar cost structure	Costs scale linearly with volume; hard ceiling on throughput; backlog and denials tend to creep back

---

Comparison Criteria

We evaluated each option against three operator-grade criteria:

• Pricing structure fit (per workflow, per volume, per site):
  How well the pricing aligns with real referral and claims volume, multiple sites, and cross-functional workflows (intake → benefits → PA → claims). Does it flex as you grow, or blow up your budget as you add locations and drugs?

• True ROI vs. FTEs (time, denials, and revenue):
  How the model compares to hiring or outsourcing in terms of minutes per document, backlog days, avoidable denials, and net revenue capture—not just “cost per license.”

• Operational burden and durability:
  How much effort it takes to deploy, maintain, and audit the automation in the messy channels where the work actually happens—payer portals, faxes, and phone calls. Does it behave like an always-on, trainable back-office team, or a set of fragile scripts?

---

How Mandolin vs. Thoughtful AI Typically Charge

Before we rank them, it helps to map the pricing “shape” of each option in the specialty-drug context.

Mandolin: AI agents priced like a back office, not a point tool

Mandolin isn’t sold as “bots per workflow.” It’s more like standing up a back office full of your best employees—priced against the volume and complexity of the work they take over:

• Pricing style:

  • Typically volume- and scope-based, not per-click or per-bot
  • Structured around end-to-end specialty-drug workflows (intake → benefits → OOP estimate → PA → claims statusing/appeals)
  • Can be aligned by patient volume, prescription/referral volume, or program scope (e.g., infusion vs. specialty pharmacy, 340B, GPO, specific therapies)

• What you’re actually paying for:

  • AI agents that read inbound referrals, labs, and clinical notes regardless of formatting
  • Agents that log into payer portals, perform benefits investigations, interpret coverage, and assemble site-specific OOP estimates (including GPO/340B and drug acquisition cost)
  • Prior auth agents that pull medical policies, compile PA packages, and submit via portal/fax/phone
  • Claims agents that check status, interpret remits, and initiate appeals—again, through the channels payers actually require

• ROI lens:
  Mandolin’s published numbers are measured like a back office, not a software feature:

  • 24x speed increase: From ~20 minutes per document to ~3 minutes, with under-2-hour turnaround
  • Backlog elimination: 4-day Rx backlog reduced to 0 days
  • Scale to 4,500+ patients/month while refocusing the work of 13 outsourced roles
  • Implicit FTE equivalence: one Mandolin deployment can absorb dozens of FTEs worth of repetitive portal/fax work

From a pricing standpoint, this translates into “per-program capacity” vs. per-widget billing. For a multi-site infusion program, you’re buying enough Mandolin “back-office horsepower” to handle today’s volume and your growth curve—without hiring or outsourcing in lockstep.

---

Thoughtful AI: RPA-style automation priced by workflows, bots, or events

Thoughtful AI is an RPA/automation provider. While details vary by contract, what you typically see in this category looks like:

• Pricing style:

  • Per workflow / per bot / per process or a bundle of “automation units”
  • May have volume tiers based on number of transactions or executions
  • Often department-agnostic (revenue cycle, finance, HR, etc.)

• What you’re actually paying for:

  • Configuration of specific workflows (e.g., scraping data from a portal, keying into a system)
  • Bot runtime for those workflows, often limited to the exact steps you define
  • Some monitoring/maintenance, but typically your team has to spec and test every change

• ROI lens:

  • Gains tend to be measured as time saved per workflow (e.g., “5 minutes saved per task”)
  • Biggest value comes from repetitive, rules-based tasks that don’t change frequently
  • In specialty access, payer-policy volatility and portal changes can erode ROI if you’re constantly re-tuning bots

RPA-style pricing can look attractive for a single, well-bounded workflow. But in specialty drugs, the work isn’t bounded: new therapies, new payers, new medical policies, and new paperwork formats show up constantly. That means more build, more re-build, and more internal ownership than most access teams have capacity for.

---

Adding FTEs: Linear cost, limited elasticity

• Pricing style:

  • Fully loaded cost per FTE (salary + benefits + overhead), often $70K–$100K+ per year in specialty access markets
  • More volume or complexity = more FTEs or more overtime, with familiar but rigid budgeting

• What you’re actually paying for:

  • Humans reading every faxed referral, portal policy, and remit
  • Staff logging into each payer portal, doing investigations, calling payers, and keying into the EHR
  • Informal “automation” via spreadsheets and tribal knowledge

• ROI lens:

  • Predictable but linear: every 20–25% volume jump usually triggers the need for more headcount or another outsourcing contract
  • Denials and time-to-therapy often move in the wrong direction when volumes spike beyond team capacity

---

Detailed Breakdown

1. Mandolin (Best overall for high-volume specialty-drug programs)

Mandolin ranks as the top choice because its pricing aligns with the end-to-end specialty-drug workflow and the ROI is measured in FTEs displaced, backlog days eliminated, and patient volume handled—not in how many bots you bought.

What it does well:

• End-to-end workflow coverage priced like capacity, not features

  • Instead of charging for each micro-workflow (e.g., “portal login bot,” “fax parser bot”), Mandolin prices against the entire access lifecycle:
    • Intake and referral normalization
    • Benefits investigations across portals and payers
    • Out-of-pocket calculations that include site-specific fee schedules, drug acquisition costs, GPO and 340B pricing, and co-pay assistance
    • Medical policy review and prior auth prep/submission
    • Claims statusing and appeals via portal, fax, and phone
  • You’re effectively buying an AI-powered access and revenue operations team that can flex across steps, sites, and therapies.

• ROI that clearly beats stacking FTEs or outsourcing

  • 24x faster document handling (20 minutes → ~3 minutes) with under-2-hour turnaround means:
    • A team that used to cap out at X documents/day can now handle 24X with no incremental headcount.
    • Backlogs that used to sit at 3–5 days can realistically sit at 0–1 day even during growth or seasonal spikes.
  • The ability to scale to 4,500+ patients/month while refocusing 13 outsourced roles is effectively reclaiming 13 FTE-equivalents and redeploying them to higher-value work.
  • Because Mandolin is built for specialty-drug economics, it doesn’t just “reduce clicks”—it protects margin by accurately reflecting site-of-care fee schedules, drug acquisition cost, and assistance programs in every estimate.

• Compliance and traceability baked into the pricing value

  • Every agent action is logged and traceable, which matters when you’re under HIPAA, BAAs, and payer audits.
  • Instead of “the bot did something weird,” you get auditable step-by-step trails: which portal was accessed, what data was read, what was submitted, and when.

Tradeoffs & Limitations:

• Best ROI at meaningful scale and complexity
  • If you run a very small, single-site program with limited therapies and low monthly volume, you may not fully leverage Mandolin’s capacity.
  • The value multiplies as you add:
    • more sites or locations
    • more biologicals and infused therapies
    • more payers and medical policies
    • more referral sources and document formats

Decision Trigger:
Choose Mandolin if you want to replace large swaths of access/back-office FTE work across intake, benefits, PA, and claims with AI agents—and you care about measurable outcomes like 0-day backlog, 24x speed gains, and scaling to thousands of patients/month without adding staff.

---

2. Thoughtful AI (Best for broad, non-specialty RPA use)

Thoughtful AI is the strongest fit here because its RPA-style pricing works well when you have rule-based, stable workflows across multiple back-office domains—not necessarily the dynamic, policy-driven world of specialty-drug access.

What it does well:

• Flexible, workflow-based automation across departments

  • Good fit for non-clinical tasks like AP/AR, HR onboarding, finance reconciliations, and generic rev cycle steps that rarely change.
  • Pricing often maps to a per-workflow or per-bot model, which can be cost-effective when you have:
    • clearly defined process steps
    • low policy volatility
    • limited portal and fax complexity

• Modular adoption and narrower pilots

  • You can start with one or two workflows and pay for just those, instead of committing upfront to an end-to-end specialty-drug access deployment.
  • For organizations early in automation, that “toe in the water” can feel safer.

Tradeoffs & Limitations:

• Maintenance burden in payer-portal and fax-heavy workflows

  • Specialty-drug access workflows live in changing payer portals, PDFs, and fax formats. Every time a portal UI shifts or a payer tweaks a policy, a workflow can break.
  • With per-workflow/bot pricing, you may:
    • pay to build the workflow
    • then pay again (in either vendor fees or internal time) every time you update it
  • ROI erodes when more of your time is spent maintaining automations than resolving denials.

• Limited economic intelligence for specialty drugs

  • RPA can push pixels, but it doesn’t inherently understand site-specific fee schedules, acquisition cost math, GPO vs. 340B dynamics, or co-pay programs.
  • You’re still on the hook to layer that logic somewhere else—and maintain it—if you want accurate out-of-pocket figures and margin protection.

Decision Trigger:
Choose Thoughtful AI if your primary goal is to automate discrete, stable back-office tasks across many departments, and you have the internal resources to design, monitor, and update RPA workflows. It’s a fit when you want per-workflow pricing and your highest-value automations are outside the specialty-drug “messy middle.”

---

3. Adding FTEs (Best for very small, stable programs)

Adding FTEs stands out for this scenario because the cost structure is familiar, the control is high, and for low-volume or early-stage programs, you can be overpowered by enterprise automation before you really need it.

What it does well:

• Direct control and immediate adaptability

  • Humans can handle edge cases, weird policies, and one-off payer requests without a build cycle.
  • When you’re onboarding a new therapy or payer, a trained specialist can still be the fastest way to learn the workflow.

• Simple, predictable budgeting at small scale

  • At 200–300 patients/month or fewer, a couple of cross-trained access specialists may be more than enough.
  • Your finance team understands the cost of one more FTE better than a new automation line item.

Tradeoffs & Limitations:

• Linear cost with a hard throughput ceiling

  • Every additional 25–30 referrals a day typically translates into either:
    • more FTEs, or
    • more overtime, burnout, and mistakes
  • Backlogs (days from fax to EHR), avoidable denials, and delayed start-of-care are the first symptoms when you hit that ceiling.

• Difficult to justify 24/7 coverage or surge capacity

  • Patients and referring sites don’t wait for your hiring plan. When volume spikes, your staff can’t magically run a 24/7 operation the way AI agents can.
  • Seasonal fluctuations and new-site go-lives tend to break the model, forcing temporary outsourcing or delays.

Decision Trigger:
Stick with or add FTEs if you are a single-site, low-volume program with stable payer mix and slow growth, and you can accept some backlog and longer time-to-therapy. Once you start planning for multi-site scale or high-growth therapy lines, FTE-only models rarely hold up economically.

---

Pricing Models vs. ROI: How the Math Typically Shakes Out

Let’s put the three options into a specialty-drug operator’s mental spreadsheet.

Assume:

• You’re running infusion + specialty-drug programs across 2–3 sites
• You handle 1,500–3,000+ patients/month, with referrals arriving in every format
• Your current process includes:
  • Fax-to-EHR intake and data entry
  • Benefits investigation via multiple payer portals and calls
  • Manual out-of-pocket estimates
  • PA compilation and submission via portal/fax/phone
  • Claims statusing and appeals

Mandolin-style pricing vs. FTEs

• Inputs:

  • Pre-Mandolin: 20 minutes/document average, multi-day backlog
  • Post-Mandolin: ~3 minutes/document, under-2-hour turnaround, 0-day backlog reported

• FTE equivalent:

  • One FTE handling 20 minutes/document can process ~24 docs/day (8 hours).
  • Mandolin at ~3 minutes/document equates to ~8x documents per “FTE-hour.”
  • When you scale to thousands of patients/month, this stacks into dozens of FTE-equivalents worth of capacity.

• Economic reality:

  • You typically replace the need to hire or outsource multiple FTEs as volume grows.
  • You also avoid revenue leakage from:
    • missed assistance opportunities
    • inaccurate OOP estimates that deter starts
    • preventable denials from mismatched policies or missing clinicals
  • Because Mandolin’s pricing is based on workflow scope and volume, you’re buying elastic capacity rather than incremental headcount.

Thoughtful-style RPA pricing vs. FTEs

• Inputs:

  • Assume you automate a slice: e.g., pulling eligibility from a portal or keying specific fields from a PDF into your system.
  • Bots can save 5–10 minutes per task, but require workflow definition, testing, and maintenance.

• FTE equivalent:

  • You may reclaim fractions of FTEs across multiple workflows.
  • In practice, some of that reclaimed time is burned back on:
    • troubleshooting broken scripts
    • updating bots when payers redesign portals or forms
    • managing exceptions that weren’t coded into the workflow

• Economic reality:

  • RPA can be cheaper than a full FTE for a well-defined, stable slice of work.
  • In specialty access, the instability (new therapies, policy shifts) can blunt the ROI unless you invest heavily in ongoing automation management.

FTE-only model: steady-state cost, unstable outcomes at scale

• Inputs:

  • Every 1–2 new referral sources or therapy lines typically requires more training and eventually more headcount.
  • A bad month (staff turnover + volume spike) often triggers backlog explosions.

• Economic reality:

  • Technically simple to model—FTE cost is known—but you increasingly lose revenue through:
    • missed or late benefits checks
    • under-optimized OOP estimates that miss assistance and 340B/GPO advantages
    • denials from rushed or incomplete PA submissions
  • At scale, “add FTEs” becomes more expensive and less effective than AI agents that never get tired, don’t churn, and can run in parallel indefinitely.

---

Final Verdict

If your world is specialty drugs—infusion, buy-and-bill, or specialty pharmacy—Mandolin’s pricing model lines up better with how the work actually happens and how you measure success. It’s effectively priced as an always-on, AI-powered back office that:

• absorbs large amounts of portal/fax/phone work
• delivers documented gains like 24x speed, 0-day backlogs, and 4,500+ patients/month capacity
• replaces the need to keep adding FTEs or outsourcing to maintain service levels

Thoughtful AI and similar RPA vendors fit best when your highest-value automations are stable, rules-based workflows across many departments and you’re comfortable paying per workflow or per bot. They can absolutely save time, but in the messy, policy-driven middle of specialty-drug operations, their per-workflow economics and maintenance burden often struggle to beat a purpose-built agentic system.

Adding more FTEs remains the simplest answer for very small or early-stage programs. But once you’re feeling the pain of backlogs, denials, and growth-driven hiring, the ROI tipping point typically favors an AI back office like Mandolin—especially when you measure in the metrics that actually matter: time-to-therapy, denials avoided, and revenue captured per patient.

---

Next Step

Get Started