I found a perfect example of the latter on a Reddit forum for therapists. The post was titled: "Medicaid billing literally makes me cry."

The author is a solo practitioner who is owed over $2,000 for services she has already provided. She submitted the same claim over 20 times, and each time it was denied with a short, cryptic explanation like "incorrect billing code." She called Medicaid for help; they told her they couldn't offer feedback. She looked for trainings and felt "even more lost."

This isn't just an administrative headache. This is a small business owner on the brink, unable to get paid for her work.

The comments section revealed this was not an isolated incident. It was a universal cry of pain.

"Medicaid in my state owes me 12k. Two months ago they owed me over 20k."

"I'm at $40k owed by Medicaid. I barely paid myself last year to make sure my employees could get paid in full and on time. It's not sustainable to live this way."

The Black Box Problem

What's really going on here? The core of the problem is that Medicaid billing is a "black box."

To get paid, a therapist has to submit a claim with a perfect combination of codes: a CPT code for the service, a diagnosis code, a "place of service" code, and often a series of "modifiers" that are specific to the state, the payor, and the context (like telehealth). If even one of these is wrong, the system automatically denies the claim.

The problem is, the system doesn't tell you what you did wrong. The denial codes are so vague that they're useless. As the original poster found, calling for help is a dead end.

This creates a brutal, financially devastating cycle of guess-and-check. Therapists, who are trained to help people with trauma, are forced to become amateur forensic accountants, trying to decipher an arcane and unforgiving system.

The "Solutions" (and Why They're Not Really Solutions)

The overwhelming consensus in the thread points to one primary solution: give up and hire a human.

"Hire a biller, Mine takes 6%, pre-certifies, fights when claims are denied, SOOOOOO worth it."

Think about that for a moment. The system is so fundamentally broken that the best-known solution is to pay someone 6% of your top-line revenue just to navigate it for you. For a solo practitioner, that is a massive and often unsustainable cost.

Other well-meaning commenters tried to help by offering their own manual checklists:

• "Did I use the right modifiers? 95 for telehealth and AJ for masters levels..."

• "Was my place of service correct? For telehealth make sure to use 10 not 02."

• "Did I do a clean claim vs an edited claim?"

While helpful, this just highlights the sheer complexity of the problem. A therapist now needs to become an expert in the subtle differences between dozens of state-specific rules and codes.

The result? Many therapists who want to serve this vulnerable population simply can't afford to. As one commenter put it: "I would love to accept Medicaid... Unfortunately, it'll never happen as the billing (and payment) makes it so complicated."

The Path Forward

This is a classic information asymmetry problem. The knowledge required to get paid is arcane, fragmented, and locked away in dense provider manuals or the heads of expensive specialists.

A system that could act as an expert co-pilot, guiding a therapist to the right combination of codes before they submit a claim, seems like a logical solution.

But what would that product actually look like? What's the specific blueprint for an MVP that could win in this space? And what are the hidden risks that explain why this opportunity still exists?

Next Tuesday’s premium brief, "Medicaid Billing Assistance for Solo Therapists," is the complete analysis of that exact question. It's a full product teardown of the space, including:

• A detailed breakdown of the competitive landscape.

• A specific, buildable product blueprint for an "AI Medicaid Co-pilot."

• A validation plan you could execute in 30 days.

Upgrade to get the full brief →

Every week, I research real problems from professional communities and analyze them as potential SaaS opportunities. Join the many builders and investors who read SaaS Fieldwork for startup ideas sourced from real workflows.

P.S. If you're a therapist who has battled the Medicaid beast, I'd love to hear your story. The most painful submission gets a free annual subscription. Misery loves company, and I love market research.

The answers were staggering: 400 in the first 24 hours. 1,200 in a few days. 1,800 before the posting was shut down.

These numbers aren't a sign of a healthy market. They are a signal of a broken system. The "why" behind this flood was perfectly articulated in a separate thread from another recruiter who had hit their breaking point. The title was: "Increasing number of spam applicants from job boards."

The comments sections of both threads painted the same picture of a system in chaos.

"Yep, job boards are a junkyard right now. I’ve seen the same resume template show up 12 times in one morning, same typos, same fake certs, all from random IPs."

"1000+ [applications] and like 95% are shit, without even having must have skills."

"Our ATS is awful so it’s manual trying to work through... IT reqs kill my team."

The AI Arms Race

What's happening here is a classic technological arms race. For years, the conventional wisdom for job seekers was to tailor their resume for every application to get past the Applicant Tracking System (ATS) keyword filters. This was a tedious, manual process.

Now, AI tools make it frictionless. A candidate can generate a hundred bespoke resumes in an afternoon and auto-apply to every relevant job posting in the country. From the candidate's perspective, in a tough market, this "spray and pray" approach is a logical strategy to maximize their odds.

But for the recruiter on the receiving end, the result is a massive signal-to-noise problem. Their primary channel for finding candidates—the public job board—has been rendered almost unusable. Their job has shifted from "evaluating talent" to "detecting spam."

The Flawed Defenses

The threads also revealed the current, often inadequate, workarounds recruiters are forced to use.

1. Manual Triage: This is the default. Recruiters spend hours sifting through resumes, looking for subtle red flags. As one user detailed, they've become forensic investigators, checking for "no linkedin, broken LinkedIn," "VOIP phone numbers," and "inconsistent or illogical information between resume and LinkedIn." This is a brutal, soul-crushing use of a skilled professional's time.

2. Abandoning the Channel: A common piece of advice was to give up on job boards entirely and shift to manual, outbound sourcing on LinkedIn or relying on referrals. This is a valid strategy, but as one of the original posters pointed out, it's not a viable solution when you need to fill high-volume roles.

This isn't a story about recruiters being lazy. It's a story about a system being overwhelmed. Their existing tools, the traditional ATS platforms, were designed for a different era. They are a castle wall built for arrows, and the enemy has just invented gunpowder.

The Opportunity in the Chaos

A system that could act as an "AI firewall," sitting in front of the ATS to filter out the 95% of noise so recruiters can focus on the 5% of signal, seems like an obvious next step.

But what would that product actually look like? What's the specific blueprint for an MVP that could win in this crowded and complex space? And what are the hidden risks—especially around AI bias and trust—that explain why this opportunity still exists?

Next week's premium brief, "Builder Brief: From 1,000 Applications to 50 Real Ones - The AI Screening Opportunity," is the complete analysis of those exact questions. It's a full product teardown of the space, including:

• A detailed breakdown of the competitive landscape.

• A specific, buildable product blueprint for an "AI Firewall."

• A validation plan you could execute in 30 days.

Upgrade to get the full brief →

Every week, I research real problems from professional communities and analyze them as potential SaaS opportunities. Join the many builders and investors who read SaaS Fieldwork for startup ideas sourced from real workflows.

P.S. If you're a recruiter who has waded through a junkyard of applications, I'd love to hear your story. The most painful submission gets a free annual subscription. Misery loves company, and I love market research.