Ogretape's Two-Stage AI Pipeline Finds 13 SaaS Ideas for Auto Detailers

Ogretape's unfairgaps-os project uses a novel two-stage AI pipeline to infer specific, buildable SaaS ideas for niche professions, generating 13 distinct tools for auto detailers alone.

Ogretape's open-source project, unfairgaps-os, processed 130 US profession profiles, seeding 25 with generated SaaS ideas. For auto detailers, the system identified 13 distinct, buildable tools, including a pricing calculator specced as a $19/month SaaS. This approach shifts product discovery from direct user interviews to a structured, AI-driven inference of pain points based on regulatory and routine data.

Two-stage pain inference

The initial version of unfairgaps-os mined court filings and regulatory data to identify B2B pain points, focusing on industry-level problems documented in lawsuits. This approach proved ineffective for individual professionals. Lawyers are not sued over tedious filing fee calculations, nor are accountants fined for annoying trust account reconciliations. The daily, recurring pains of a working professional rarely appear in court records.

Ogretape pivoted to a two-stage pipeline specifically for individual professionals. Stage 1 employs WebSearch with seven targeted queries. These queries pull regulatory facts from authoritative sources such as .gov domains, law.cornell.edu, the Bureau of Labor Statistics (BLS), and professional association websites. The output is a structured JSON profile containing approximately 30 specific facts and their source URLs for each profession. This data covers daily routines, documents, regulations, licensing requirements, common software, jargon, career levels, fears, professional communities, and labor market information.

Stage 2 takes this structured JSON profile and feeds it to Opus 4.7, an LLM. Crucially, the LLM operates with a deductive prompt and no web access, preventing external data contamination. Given the detailed regulatory and routine context, the LLM infers 8 to 15 specific, recurring tasks that would be painful for the professional. For each identified pain point, the system produces a structured specification for an AI tool designed to solve it.

Structured output for SaaS ideas

The output for each inferred pain point is a detailed, buildable specification, not a general platitude. For auto detailers in the US, the system generated 13 distinct tool ideas. These ranged from a calculator to price detail jobs profitably to a checklist for EPA stormwater compliance.

One example is the pricing calculator. Most detailers, according to the system's inference, eyeball pricing and often undercut their services by 25% due to a lack of a real cost-plus formula. The output JSON for this tool includes the precise formula: labor, chemicals, the 2026 IRS $0.67/mile rate, 15.3% self-employment tax, and monthly overhead allocation. It also specifies 10 inputs, including a services list and target margin, and outputs like minimum profitable price, recommended price with margin, a breakdown, and tax set-aside. This forms a complete specification for a $19/month SaaS product.

Another example is the EPA stormwater compliance checklist. The system highlights the severe risk: a Clean Water Act civil penalty of $64,618 per day per violation for dumping wash water into storm drains. The output provides a 12-step compliance procedure, including warnings that biodegradable soap is not a defense, and legal citations such as 33 USC 1311 and 40 CFR 122.26. Each of the 13 identified tools received a similarly structured and actionable specification.

Data sources and targets

The efficacy of Ogretape's pipeline relies on the quality and specificity of its initial data collection. The WebSearch stage targets highly authoritative and domain-specific sources. These include government websites (.gov), legal databases (law.cornell.edu), labor statistics (BLS), and professional association sites. This ensures the regulatory and routine information is accurate and relevant to the profession being analyzed.

The seven targeted queries cover a comprehensive range of professional life aspects. They gather data on daily routines and required documents, specific regulations and licensing processes, commonly used software, industry jargon, typical career levels and associated fears, relevant professional communities, and current labor market conditions. This broad data intake allows the LLM in Stage 2 to perform a robust deductive analysis, inferring pains that arise from the intersection of regulatory mandates and practical daily operations.

Quantifying the opportunity

The system's ability to generate detailed, actionable specs for potential SaaS products provides a quantifiable starting point for founders. The auto detailer pricing calculator, for instance, is not merely an idea but a fully outlined product with a suggested price point of $19/month, complete with its underlying financial logic and required inputs. This level of detail reduces the initial guesswork typically involved in product ideation.

The identification of high-stakes compliance issues, such as the EPA Clean Water Act violations, underscores another dimension of opportunity. The potential $64,618 per day penalty for improper wash water disposal highlights a critical pain point where professionals would likely pay for a reliable, legally cited compliance solution. By providing specific legal references like 33 USC 1311 and 40 CFR 122.26, the system offers a foundation for building tools that address significant regulatory risks.

Ogretape loaded 130 US profession profiles into the repository and ran Stage 2 on 25 of them to seed the system. This indicates a scalable method for generating a substantial volume of potential SaaS ideas across diverse professional niches, moving beyond anecdotal problem discovery.

WHAT WE'D CHANGE

Ogretape explicitly states this system is a

Pull quote: “”