Product Strategy

Chapter 13 — Product Strategy

RAPID AI: From Domain Expertise to Scalable Platform

“The world does not need another vibration dashboard. It needs a system that thinks like a reliability engineer — and scales like software.”

13.1 The Three-Track Business Model

RAPID AI’s product architecture is not an accident of software design. It mirrors the three tracks of reliability engineering that Dibyendu De refined over 28 years of field diagnostics. Each track represents a deeper level of engineering intelligence, a higher barrier to entry for competitors, and a correspondingly higher return on investment for customers.

Track 1: AI-Enabled Prescriptive CBM — The Entry Point

What it does: Real-time condition monitoring with physics-based diagnostics and prescriptive maintenance actions. Modules A through E of the platform: signal validation, fault detection, system stability assessment, health staging, and maintenance action planning.

ROI: 1:10 or greater.

This is the SaaS entry point. A plant connects its vibration and process sensors to RAPID AI. Within minutes, the system validates signal quality (Module A), detects fault patterns across 119 physics-based rules and 12 component types (Module B), fuses the evidence into a system stability index (Module C), stages health and estimates remaining useful life (Module D), and prescribes specific maintenance actions with priority rankings and justification chains (Module E).

The critical differentiator at Track 1 is the last mile. Every competitor can detect an anomaly. RAPID AI explains why the anomaly exists, traces it to a specific failure mechanism, and prescribes what to do about it — with every recommendation linked back to the physics that triggered it. An engineer does not receive an opaque “anomaly score of 0.87.” They receive: “Bearing outer race defect detected (confidence 78%), driven by lubricant film breakdown (initiator), accelerated by axial preload above design spec (accelerator). Recommended action: replace lubricant with ISO VG 68 equivalent, verify axial clearance, schedule bearing inspection within 14 days.”

Pricing model: Monthly subscription per monitored asset. Tiered by asset count (50, 200, 500, 1000+) with volume discounts. Entry price point: $150-300/asset/month depending on industry and asset criticality.

Target customer: Plants that already have sensors and basic CBM programs but are drowning in alarms they cannot interpret. They want to move from “we know something is wrong” to “we know what is wrong, why it is wrong, and what to do about it.”

Track 2: AI-Driven RCM Optimization — The Strategic Tier

What it does: Dynamic Reliability-Centered Maintenance workbooks where risk scores update automatically based on real-time condition data. Maintenance strategies shift in real time as machine health changes. Includes all of Track 1 plus Modules F (Weibull reliability) and enhanced RCM decision logic.

ROI: 1:50 or greater.

Traditional RCM is a periodic exercise — a team of engineers spends weeks reviewing failure modes, assigning risk rankings, and selecting maintenance strategies. The resulting workbook is static. It reflects the state of the plant at the time of the review. Six months later, operating conditions have changed, new failure modes have emerged, and the workbook is already stale.

RAPID AI’s Track 2 makes RCM dynamic. Every failure mode in the workbook carries a live risk score computed from real-time condition data, Weibull reliability parameters adjusted for actual operating conditions, and the IAR (Initiator-Accelerator-Retarder) classification that determines whether the failure pathway is being contained or accelerating. When a bearing’s condition shifts from “degrading” to “warning,” the RCM workbook automatically re-ranks that failure mode, adjusts the maintenance interval, and escalates the strategy from “monitor” to “plan intervention.”

Pricing model: Enterprise annual license. Includes Track 1 capability for all monitored assets plus the RCM optimization module. Price point: $200K-500K/year depending on plant size and asset count, with implementation services included.

Target customer: Reliability engineering teams that conduct formal RCM reviews (typically in regulated industries — oil and gas, power generation, nuclear). They already understand the value of RCM but are frustrated by the static nature of traditional workbooks and the labor intensity of periodic reviews.

Track 3: RAPID Innovation / Design-Out — The Premium Tier

What it does: Contradiction-Driven Engineering (CDE) for chronic, high-cost failures that maintenance alone cannot solve. Module G of the platform: identifying the engineering contradictions that make a failure chronic, and recommending design modifications that eliminate the root cause permanently.

ROI: 1:100 or greater.

Some failures are not maintenance problems. They are design problems. A pump that fails every 14 months due to seal contamination does not need better seal replacement procedures. It needs a seal redesign, or a process modification that eliminates the contaminant, or an operating protocol change that keeps the pump out of the conditions that cause seal degradation. Track 3 identifies these contradictions systematically.

The CDE module applies 8 contradiction types and 7 resolution patterns drawn from Dibyendu De’s diagnostic framework. When the system detects a chronic failure — one that recurs despite correct maintenance execution — it maps the engineering contradiction: “Increasing seal pressure improves containment but accelerates wear.” It then searches the resolution space for design modifications that resolve the contradiction without introducing new failure modes.

Pricing model: Consulting-augmented engagement. Annual retainer ($300K-1M+) that includes platform access, dedicated domain engineering support, and quarterly design review sessions. This tier requires human expertise — the platform identifies the contradictions, but resolving them requires engineering judgment that cannot be fully automated today.

Target customer: Plants with chronic, high-cost failures — the failures that consume 80% of the maintenance budget despite representing 20% of the failure modes. Typically large refineries, power generation complexes, and mining operations where a single chronic failure can cost $2-5M per year in unplanned downtime, collateral damage, and lost production.

13.2 Market Positioning

The Competitive Landscape

The industrial predictive maintenance market is crowded with well-funded incumbents and well-hyped startups. Understanding what they do well — and where they fall short — is essential to positioning RAPID AI.

Competitor	Strength	Weakness
Emerson (AMS)	Deep installed base, sensor hardware integration	Rules are threshold-based; no physics reasoning; vendor lock-in
SKF (Enlight)	Bearing domain expertise, vibration sensors	Narrow component focus; no system-level stability analysis
Augury	Modern UX, easy onboarding, ML-driven	Black-box ML; no explainability; limited to pattern matching
Senseye (Siemens)	Enterprise integration, Siemens ecosystem	Statistical anomaly detection only; no root cause; no prescriptive actions
SparkCognition	Strong ML/AI brand, government contracts	Domain-agnostic; requires extensive training data; no physics foundation
Uptake	Fleet-level analytics, good visualization	Correlation-based; struggles with novel failure modes; data-hungry
Aveva (Schneider)	Process integration, historian connectivity	Focused on process optimization, not machine diagnostics

RAPID AI’s Differentiation

Four structural advantages separate RAPID AI from every competitor in the market:

1. Physics-based, not black-box ML. Every diagnostic rule in RAPID AI is grounded in mechanical physics — energy flow through machine elements, thermodynamic principles, material science. The system does not need thousands of historical failure examples to recognize a bearing outer race defect. It recognizes the defect because it understands the physics of how defect frequencies manifest in vibration spectra, how lubrication film breakdown alters the energy dissipation path, and how temperature rise correlates with friction-induced energy conversion. This means RAPID AI works on day one, on any machine, without training data.

2. Explainable, not opaque. Every recommendation traces to specific rules, specific measurements, and specific physics. The justification chain is transparent: “Rule AFB_OR_001 triggered because envelope vibration at BPFO frequency exceeded 2.5g (threshold: 1.8g), corroborated by temperature rise of 3.2 degrees C/week (Rule AFB_TEMP_003).” An engineer can follow the reasoning, agree or disagree, and build trust incrementally. This is not a feature — it is a regulatory requirement in safety-critical industries.

3. Comprehensive rule coverage. RAPID AI’s rule system contains 451+ physics-based rules across four layers (Guard, Sense, Fuse, Act), covering 12 component types. A typical competitor’s system has 10-20 alarm thresholds per component type. The depth of coverage means RAPID AI can distinguish between failure modes that look similar in raw vibration data but have fundamentally different root causes — and therefore require fundamentally different corrective actions.

4. Design-out capability. No competitor in the market offers Module G — Contradiction-Driven Engineering. Every other system stops at “here is what is wrong” or, at best, “here is what maintenance action to take.” RAPID AI goes further: “here is why this failure is chronic, here is the engineering contradiction that makes it recur, and here is how to redesign the system so it never fails this way again.” This capability has no equivalent in the market.

The “Last Mile” Problem

The industrial maintenance industry has a well-known gap that existing tools have failed to close. Current systems can detect that something is abnormal. Some can classify the anomaly into a category. Almost none can answer the three questions that matter:

WHY is it failing? (Root cause mechanism, not just pattern label)
HOW FAST is it getting worse? (Trajectory and remaining useful life, not just current severity)
WHAT EXACTLY should we do? (Specific action with justification, not just “schedule maintenance”)

This is the last mile of predictive maintenance — the gap between detection and decision. RAPID AI exists to bridge it. The platform converts raw sensor data into engineering decisions: what is failing, why it is failing, how fast it is progressing, what to do about it, and whether the failure can be permanently eliminated through design modification.

13.3 Target Industries and Customers

Primary Industries

Oil and Gas (Refineries, Pipelines, Offshore Platforms)

Highest concentration of rotating equipment per site
Regulatory pressure (API, OSHA, EPA) demands documented risk management
Unplanned downtime costs: $500K-2M per day per process unit
Established CBM programs with massive data but poor diagnostic utilization
Typical asset count: 2,000-10,000 rotating machines per refinery

Power Generation (Turbines, Generators, Cooling Systems)

Critical assets with high consequence of failure (grid stability, safety)
Long asset lifecycles (30-50 years) with evolving failure modes
Regulated maintenance requirements (NERC, NRC for nuclear)
Strong existing vibration monitoring programs — perfect upgrade path
Typical asset count: 200-1,000 critical rotating machines per plant

Mining and Minerals (SAG Mills, Crushers, Conveyors)

Extreme operating conditions accelerate wear mechanisms
Remote locations make unplanned maintenance extremely costly
High asset criticality — a SAG mill failure idles the entire processing circuit
Less mature CBM programs — larger opportunity for greenfield deployment
Typical asset count: 500-3,000 rotating machines per mine site

Secondary Industries

Cement and Heavy Manufacturing — High-severity operating conditions, dust contamination, thermal cycling. Plants are cost-sensitive but failure costs are high. Good fit for Track 1.

Petrochemicals — Similar profile to oil and gas but with additional chemical compatibility constraints. Strong regulatory drivers. Good fit for Tracks 1 and 2.

Steel and Metals — Extreme temperatures, heavy loads, continuous operation. Rolling mills, blast furnace auxiliaries, and continuous casters have well-documented failure modes that map directly to RAPID AI’s rule system.

Pulp and Paper — Large rotating equipment fleets, moisture-related failure mechanisms, cost-driven industry. Price-sensitive but high asset count per site creates volume opportunity.

Target Personas

Persona	Pain Point	RAPID AI Value
Reliability Engineer	Drowning in alarm data, cannot diagnose root causes fast enough	Physics-based diagnosis in minutes, not days
Maintenance Manager	Cannot prioritize work orders by actual risk	Risk-ranked prescriptive actions with justification
Plant Manager	Unplanned downtime eroding production targets	Remaining useful life estimates enable planned intervention
VP Operations	Maintenance costs rising without reliability improvement	Track 2 RCM optimization; Track 3 chronic failure elimination
Chief Engineer	Chronic failures that maintenance cannot solve	CDE contradiction analysis and design-out recommendations

13.4 Go-to-Market Strategy

Phase 1: Prove It (Months 1-12)

Objective: Validate the platform with 2-3 anchor customers. Demonstrate measurable ROI. Build reference cases.

Identify 2-3 industrial plants with existing vibration monitoring infrastructure (sensor data already available — reduces deployment friction)
Target plants with 50-100 critical rotating assets each
Deploy Track 1 (Prescriptive CBM) on a subset of assets (10-20 per plant)
Measure: Mean Time to Diagnosis, diagnostic accuracy vs. manual expert assessment, false positive rate, customer-reported ROI
Goal: Achieve 90%+ diagnostic accuracy and 10:1 ROI within 6 months per pilot
Pricing during pilot: Reduced rate or success-based pricing to lower adoption barrier

Key risk: Enterprise sales cycles in heavy industry are 6-18 months. The pilot itself may take 3-6 months to produce statistically meaningful results. Total time from first contact to reference case: 12-18 months.

Mitigation: Target plants where the decision-maker is the reliability engineering team lead (faster decision cycle) rather than corporate procurement (slower). Offer to run RAPID AI in parallel with existing systems — no rip-and-replace required.

Phase 2: Verticalize (Months 12-24)

Objective: Expand from general-purpose platform to industry-specific solutions with pre-configured rule packs.

Package industry-specific rule configurations: “RAPID AI for Refineries,” “RAPID AI for Power Generation,” “RAPID AI for Mining”
Each vertical pack includes: pre-configured component types, industry-standard alarm thresholds (API 670, ISO 20816, etc.), industry-specific failure mode libraries, regulatory compliance reporting templates
Expand Track 1 deployments to full plant scale (200-1,000 assets per customer)
Introduce Track 2 (RCM Optimization) to anchor customers who have validated Track 1
Begin building partner channel: vibration analysis service providers who resell RAPID AI as their diagnostic engine

Phase 3: Platform Play (Months 24-36+)

Objective: Transform from a product company into a platform company.

Open a marketplace where domain experts can publish and monetize rule packs for specific equipment types, failure modes, or industries
Enable third-party sensor integrations through a documented API
Introduce a “community” tier for independent vibration analysts and small consulting firms
Build network effects: more assets monitored produces more validated diagnostic cases, which improves rule accuracy, which attracts more customers

13.5 Revenue Model

Revenue Streams

Stream	Tier	Model	Indicative Pricing
Prescriptive CBM	Track 1	Per-asset monthly subscription	$150-300/asset/month
RCM Optimization	Track 2	Enterprise annual license	$200K-500K/year
Design-Out Consulting	Track 3	Annual retainer + project fees	$300K-1M+/year
Rule Pack Marketplace	Future	Revenue share (70/30 creator/platform)	Variable
API Access	Future	Usage-based (per analysis call)	$0.50-2.00/call

Unit Economics (Track 1 Illustrative)

A mid-size refinery with 500 monitored assets at $200/asset/month generates $100K/month ($1.2M/year) in recurring revenue. Marginal cost of serving an additional asset is near-zero once the platform is deployed — the physics engine processes each reading in under 50ms. Gross margins at scale should exceed 80%.

Revenue Trajectory (Conservative)

Year	Anchor Customers	Total Assets	ARR (Track 1)	ARR (Track 2+3)	Total ARR
1	2-3	150-300	$360K-720K	$0	$360K-720K
2	5-8	1,000-2,000	$2.4M-4.8M	$400K-1M	$2.8M-5.8M
3	10-15	3,000-5,000	$7.2M-12M	$2M-5M	$9.2M-17M

These projections assume no marketplace revenue and no API revenue. Both represent meaningful upside if the platform play succeeds.

13.6 Moat and Defensibility

Sustainable competitive advantage in industrial AI does not come from algorithms. Algorithms are commoditized. It comes from domain knowledge that is difficult, expensive, and time-consuming to replicate.

The Four Layers of the Moat

Layer 1: 451+ Physics-Based Rules These rules encode 28 years of field diagnostic expertise across 50+ tier-one industrial clients and 4,000+ validated diagnostic cases. They are not statistical patterns extracted from data. They are physics-first engineering rules — each one grounded in mechanical principles, validated against real-world failures, and refined through decades of practice. A competitor starting from scratch would need to either hire a domain expert with equivalent experience (there are very few in the world) or accumulate equivalent field data over many years. Neither path is fast.

Layer 2: The Integrated Master Schema (IMS) The IMS is the ground-truth matrix that connects failure modes, diagnostic signatures, causal factors, maintenance actions, and design modifications into a single relational structure. It contains 4,000+ validated diagnostic cases mapped across 34 columns per record. This is not a database — it is a validated knowledge graph of how industrial machines actually fail and how those failures are correctly diagnosed and resolved. No competitor has an equivalent structure because no competitor has the field diagnostic corpus to populate it.

Layer 3: The FRETTLSM Taxonomy FRETTLSM (Friction, Resonance, Electrical, Thermal, Transient, Lubrication, Structural, Material) is an 88-factor causal taxonomy unique to Dibyendu De’s diagnostic framework. It provides the systematic vocabulary for classifying why machines fail — not just what fails, but what physical mechanism caused the failure, at what level of the causal chain. This taxonomy is the intellectual scaffolding of the entire rule system. It cannot be reverse-engineered from the platform’s outputs because the taxonomy organizes how rules are constructed, not just what they detect.

Layer 4: Network Effects (Future) As more assets are monitored by RAPID AI, the system accumulates more validated diagnostic cases. Each confirmed diagnosis — where an engineer validates the platform’s recommendation against actual field findings — strengthens the evidence base for that rule. Over time, this creates a flywheel: more assets produce more validated cases, which produce more accurate rules, which attract more customers, which produce more validated cases. This network effect does not exist today but becomes the dominant moat at scale.

What Is Not a Moat

The technology stack (Python, FastAPI, SvelteKit, PostgreSQL) is not a moat. The ML models used for signal processing are not a moat. The dashboard UX is not a moat. All of these can be replicated by a well-funded competitor in months. The moat is the domain knowledge encoded in the rules, the validated case library, and the causal taxonomy. These require years of field experience, not engineering sprints.

13.7 Success Metrics and KPIs

Diagnostic Performance

Metric	Definition	Target	Benchmark
Mean Time to Diagnosis (MTTD)	Time from data ingestion to actionable diagnosis	< 5 minutes	Dibyendu achieves ~30 min manually; typical analyst takes 2-4 hours
Diagnostic Accuracy	% of diagnoses confirmed correct by field inspection	95%+	Dibyendu achieves 98% manually across 4,000+ cases
False Positive Rate	% of alerts that do not correspond to real faults	< 5%	Industry average for threshold-based systems: 30-50%
False Negative Rate	% of real faults not detected by the system	< 2%	This is the safety-critical metric — a missed fault can cause catastrophic failure
Diagnostic Depth	% of diagnoses that include root cause, not just fault label	> 90%	Most competitor systems: < 10%

Business Performance

Metric	Definition	Target (Year 1)	Target (Year 3)
Customer ROI	Documented savings / platform cost	10:1	20:1
Net Revenue Retention	Revenue from existing customers year-over-year	> 120%	> 140%
Time to Value	Days from deployment to first actionable diagnosis	< 7 days	< 1 day
Asset Expansion Rate	% increase in monitored assets per customer per year	50%	100%
Pipeline Latency	End-to-end processing time per reading	< 50 ms	< 100 ms (at 10x scale)

Platform Health

Metric	Definition	Target
Rule Coverage	Active rules / total documented rules	> 90% (currently 332/454)
Test Coverage	Automated test coverage of physics engine	> 85% (currently ~80%)
System Uptime	Platform availability	99.9%
Data Quality Pass Rate	% of incoming sensor data passing Module A guard rules	> 95%

13.8 Risk Factors

Honest assessment of the risks that could slow or derail the product strategy. Each risk includes severity, likelihood, and the mitigation approach.

Risk 1: Sensor Data Quality Variance

Severity: High. Likelihood: Certain.

Sensor data quality varies wildly across industrial plants. Some plants have modern, well-calibrated vibration monitoring systems with historian databases and consistent sampling rates. Others have ad-hoc sensor installations, inconsistent measurement practices, and data gaps measured in months. RAPID AI’s physics engine requires minimum data quality to produce reliable diagnoses — garbage in still produces garbage out, no matter how sophisticated the rules.

Mitigation: Module A’s 16 guard rules (DG001-DG019) provide a hard quality gate. The system refuses to diagnose when data quality is below threshold, rather than producing unreliable results. Data quality scores propagate through the entire pipeline, degrading confidence rather than hiding uncertainty. The onboarding process includes a data quality assessment that identifies gaps before the customer expects results.

Risk 2: Domain Expertise Concentration

Severity: Critical. Likelihood: Current reality.

Twenty-eight years of diagnostic expertise currently reside primarily in one person: Dibyendu De. If this expertise is not successfully encoded into the platform’s rule system, knowledge graph, and documentation, the company’s core asset walks out the door every evening.

Mitigation: This book — and the entire RAPID AI platform — is the mitigation. The 451+ rules, the IMS case library, the FRETTLSM taxonomy, and the CDE contradiction framework are the systematic extraction and encoding of Dibyendu’s expertise into computational form. The process is ongoing: every validated diagnostic case adds to the encoded knowledge base. The goal is not to replace the domain expert but to ensure the knowledge survives, scales, and improves beyond any single person’s capacity.

Risk 3: Enterprise Sales Cycle Length

Severity: Medium. Likelihood: Certain.

Industrial enterprises do not buy software quickly. Procurement processes involve technical evaluation, pilot programs, security reviews, IT integration assessments, and multi-level approval chains. A typical sales cycle for a new industrial software platform is 6-18 months from first contact to signed contract.

Mitigation: Reduce friction at every stage. Offer parallel deployment (run alongside existing systems, no rip-and-replace). Provide a self-service trial for Track 1 with sample data. Target the reliability engineering team lead as the initial champion — they have technical authority and can often run a pilot without full procurement involvement. Structure contracts with low entry commitment and expansion pricing that rewards success.

Risk 4: Incumbent Distribution Advantage

Severity: Medium. Likelihood: High.

Emerson, SKF, and Siemens have existing relationships with every major industrial plant in the world. They sell sensors, historians, and DCS systems. Their predictive maintenance software is often bundled with hardware purchases or sold through established service agreements. Competing for shelf space against an incumbent who already owns the sensor infrastructure is an asymmetric fight.

Mitigation: Do not compete on distribution. Compete on diagnostic depth. Position RAPID AI as the intelligence layer that sits on top of existing sensor infrastructure — not a replacement for Emerson’s sensors or SKF’s vibration analyzers, but the diagnostic reasoning engine that makes those investments actually pay off. “You already have the sensors. You already have the data. What you do not have is the engineering intelligence to turn that data into decisions. That is what RAPID AI provides.”

Risk 5: Regulatory and Safety Liability

Severity: High. Likelihood: Low but consequential.

If RAPID AI recommends an action (or fails to recommend an action) and a machine failure causes injury, environmental release, or significant property damage, the liability implications are serious. Industrial safety is a regulated domain, and any system that provides maintenance recommendations must be positioned carefully relative to the engineer’s professional judgment.

Mitigation: RAPID AI is positioned as a decision-support tool, not a decision-making tool. Every recommendation is presented as advisory, with full transparency into the evidence and reasoning. The engineer remains the decision-maker. Terms of service, product documentation, and the UI itself reinforce this positioning. The explainability of the system (every recommendation traces to specific physics) actually reduces liability relative to black-box ML systems that cannot explain their recommendations.

Risk 6: Rule System Maintenance Burden

Severity: Medium. Likelihood: Moderate.

451+ rules must be maintained, validated, and updated as new failure modes are discovered, new equipment types are onboarded, and engineering standards evolve. If the rule system becomes stale or inconsistent, diagnostic accuracy will degrade and customer trust will erode.

Mitigation: Rules are stored as versioned data (YAML and Python), not hardcoded logic. Module 15 (Learning and Governance) includes rule version control, validation tracking, and licensing enforcement. The marketplace strategy (Phase 3) distributes the maintenance burden: domain experts who publish rule packs are incentivized to keep them current because their revenue depends on adoption.

13.9 Strategic Summary

RAPID AI’s product strategy rests on a simple thesis: the industrial world has more sensor data than it knows what to do with, and the bottleneck is not data collection but engineering interpretation.

The platform addresses this bottleneck at three levels of increasing value:

Track 1 turns raw data into specific, physics-backed diagnoses and prescriptive actions. This is the entry point — high volume, recurring revenue, immediate value.
Track 2 turns diagnostic intelligence into dynamic risk management. This is the strategic tier — higher revenue per customer, deeper integration, stronger retention.
Track 3 turns accumulated diagnostic knowledge into permanent failure elimination. This is the premium tier — highest value, highest margin, hardest to replicate.

The defensibility of this strategy does not depend on proprietary algorithms or platform lock-in. It depends on something harder to build: 28 years of validated domain expertise, systematically encoded into 451+ physics-based rules, a 4,000+ case knowledge graph, and an 88-factor causal taxonomy that no competitor has and no amount of training data alone can produce.

The path forward is clear. Prove the diagnostic accuracy with anchor customers. Build the reference cases. Expand vertically. Open the platform. Let the physics do the selling.

Chapter 13 of the RAPID AI Book. Domain knowledge: Dibyendu De. Technical architecture: Rick D.

Standards Alignment

Standard	Relevance to This Chapter
ISO 55000/55001 — Asset management	The three-track business model (CBM, RCM, Design-Out) maps directly to ISO 55000’s asset lifecycle management tiers, with each product tier delivering progressively deeper asset management capability.
ISO 13374 — Condition monitoring and diagnostics of machines	The API-first product architecture ensures that all ISO 13374 processing outputs are accessible to enterprise consumers through standard REST interfaces, enabling integration with existing industrial IT ecosystems.
SAE JA1011/JA1012 — RCM evaluation criteria	Track 2 pricing and product design are built around delivering SAE JA1011-compliant dynamic RCM workbooks as a continuous service rather than a one-time consulting engagement.

Changelog

Version	Date	Author	Changes
2.1.0	2026-03-17	Rick D	Added standards alignment, living doc metadata, changelog
2.0.0	2026-03-17	Rick D	Enriched with production codebase content
1.0.0	2026-03-17	Rick D	Initial chapter creation