DAII™ Diamond Quality Index: From Tolkowsky’s Brilliant Cut Theory to Modern AI Quality Index

23 hours ago
9 min read

The Origin and Core Definition of DAII™

RAGAZZA DAII™ (Diamond Artificial Intelligence Index) is a diamond quality model that combines Marcel Tolkowsky’s brilliant cut proportions with modern AI analysis. Its main purpose is to transform scattered and complex data from diamond certificates into a single, quantifiable, and easy-to-compare quality score : The DAII™ Score.

RAGAZZA Diamond artificial intelligence index

The beauty of a round brilliant diamond relies on strict mathematical and optical principles. Factors like brightness, fire, scintillation, and visual stability are directly determined by proportions, angles, and structural balance. Tolkowsky revolutionized this by shifting diamond cutting from mere craftsmanship to scientific analysis, establishing the foundation for modern cut grading.

However, today's market is far more complex, featuring natural and lab-grown diamonds, various certificate systems, and diverse pricing tiers. Standard 4C grading and basic certificate data are no longer enough to fully represent a diamond's true quality.

RAGAZZA DAII™ solves this data complexity. Instead of repeating standard 4C grades, it integrates multiple quantifiable parameters—including proportions, angles, symmetry, fluorescence, and certificate confidence - into a structured, comparable quality model. Grounded in Tolkowsky's proportional theories and powered by modern AI data analysis, DAII™ establishes a standardized language for contemporary diamond quality.

Origin: From Tolkowsky to AI-Driven Diamond Evaluation

The theoretical foundation of the modern round brilliant cut is inseparable from Marcel Tolkowsky. Tolkowsky's importance lies not just in proposing a specific set of round diamond proportions, but in introducing a diamond evaluation method: that a diamond's beauty can be mathematically analyzed through proportions, angles, and light paths.

His core question was: What round diamond proportions can more effectively allow light to enter the diamond, reflect and refract internally, and return to the viewer's eye?

This question remains the core of round diamond evaluation today.

The visual performance of a round diamond is not determined by Color or Clarity alone, nor is it decided solely by a single Cut grade. Proportion data, including Table, Depth, Crown Angle, Pavilion Angle, Crown Height, and Pavilion Height, all affect the diamond's light return, dispersion, scintillation, and face-up appearance.

However, the modern market no longer revolves around a single center of proportions. With advancements in cutting technology, changing commercial demands, the supply of lab-grown diamonds, and evolving consumer preferences, round diamond proportions now have a wider acceptable range. For example, a larger Table may alter the face-up light distribution and visual proportions; a larger face-up diameter can increase the visual size for the same carat weight; and certain Crown/Pavilion combinations, even if not perfectly matching traditional central proportions, can still achieve a balance between brightness, fire, and diameter in actual visual performance.

Therefore, modern round diamond quality evaluation cannot rely on a single concept of "ideal proportions," nor can it depend solely on the diamond Cut grade. It requires an analytical model capable of simultaneously processing multiple parameters, ranges, weights, and varying levels of certificate confidence.

DAII™ was created based on this exact logic.

Tolkowsky provided the theoretical foundation for round diamond proportions and optical evaluation; DAII™ transforms this proportional philosophy into a structured quality index for the AI era.

In short: Tolkowsky mathematized round diamond aesthetics. DAII™ models diamond quality evaluation.

Problem: 4C and Certificates Cannot Fully Explain Diamond Differences

The 4Cs (Carat, Color, Clarity, and Cut) are the most important shared language in the diamond market, and certificates provide standard data for trading. However, the limitations of the 4Cs and certificates are clear: each grade is a range, not an exact value.

D Color is not a single color value, VS1 Clarity is not a single inclusion state, and an Excellent Cut is not a single proportional structure. Two diamonds both certified by GIA as 1.00 ct, D, VS1, Excellent can still differ in actual quality.

These differences come from deeper data on the certificate, such as Table, Depth, Crown Angle, Pavilion Angle, Polish, Symmetry, Fluorescence, Girdle, Culet, and the overall combination of proportions. Together, these parameters affect the round diamond's optical geometry and visual stability.

According to DAII™ internal model estimations, two round diamonds both graded as GIA Excellent Cut can show a model gap of about 10% to 25% in their proportion and optical geometry sub-scores if one is close to the high-stability range and the other is in the borderline Excellent range. In extreme cases within the same cut grade, the overall DAII™ Score can show a gap of about 5% to 15%.

This does not mean the visual sparkle seen by the naked eye differs by exactly 15%. Actual visual differences are affected by the environment, light sources, and the diamond's physical state. Instead, this value accurately reflects that within the DAII™ quantifiable quality model, diamonds with the same certificate grades still have clear differences in proportions and structure.

Therefore, a higher single grade does not guarantee higher overall quality. For example, a D / VS1 / Excellent round diamond with proportions deviating from the high-score DAII range may receive a lower overall quality model score than a D / VS2 / Excellent round diamond whose table, depth, crown angle, and pavilion angle combination are closer to the ideal optical structure.

This does not reject the 4Cs or certificates; they remain the foundational data for evaluation. However, they mainly answer "which grade does this diamond belong to in the four main categories?" rather than fully answering "what is the overall quality position of this diamond?"

DAII™ is built to address this evaluation gap

Philosophy: Converting Diamond Evaluation into Structured Intelligence

Traditional diamond quality evaluation relies primarily on three levels: 4C grading, certificate data, and professional expertise. All three have value, but each has limitations:

The 4Cs provide basic classification, but the data dimension is limited.
Certificates provide standardized data, but still require professional interpretation.
Professional expertise can handle details, but evaluations can vary between different assessors.

The philosophy of DAII™ is to convert this scattered data into a more consistent, comparable, and computable system of structured intelligence.

Structured intelligence does not mean simply stacking more data together. Instead, it establishes a clear logic for quality evaluation: which parameters should be included, which parameters carry higher influence, which proportion ranges offer higher stability, which factors require risk adjustments, and how different certificate systems affect final quality confidence.

Within DAII™, the 4Cs are no longer viewed as the complete answer, but as the first layer of data in the quality model. Round diamond quality is determined not only by Color, Clarity, and Cut, but is also jointly influenced by proportional geometry, cutting execution, visual risks, and certificate confidence.

This means diamond comparison should move away from linear judgments, such as "D is always better than E," "VVS1 is always better than VS1," or "Excellent always represents the best." True quality structure often comes from the combination of multiple parameters.

The role of DAII™ is to transform these combinations into a computable model output.

Framework: The Core Diamond Parameters of DAII™ Analysis

The analytical framework of DAII™ is built upon multiple quantifiable diamond data points. Instead of simply repackaging the 4Cs into a single score, it integrates data from certificates and diamond specifications into layers, forming a comparable and explainable quality model.

DAII™ utilizes a weighted index model. Its public model can be summarized as:

RAGAZZA DIAMOND Artificial Intelligence Index Formula

xᵢ represents the i diamond parameter, such as Color, Clarity, Table, Depth, or Crown Angle;
sᵢ(xᵢ) represents the sub-score of that parameter after DAII standardization;
αₖ,ᵢ represents the weight of that parameter within its respective layer;
Sₖ represents the layer score of the k quality layer;
wₖ represents the relative weight of that quality layer in the overall model;
L_cert represents the certificate confidence coefficient.

In other words, DAII™ is not a simple average, but a layered, weighted model. It first converts each piece of diamond data into a standardized sub-score, then weights it according to its importance in quality evaluation, and finally generates a single DAII™ Score

The Five Core Layers of the DAII™ Model

DAII™ Layer	Key Parameters	Features
Layer 1 Quality Specifications	Color、Clarity、Cut	Establishing a Diamond Quality Baseline
Layer 2 Cut Quality	Polish、Symmetry	Assessing Facet Finish & Geometric Stability
Layer 3 Proportions & Optical Geometry	Table、Depth、Crown Angle、Pavilion Angle、Crown Height、Pavilion Height	Analyzing Brightness, Fire, Light Return & Visual Balance
Layer 4 Structural & Visual Considerations	Fluorescence、Girdle、Culet、Ratio	Evaluating Visual Risks, Durability & Outline Stability
Layer 5 Certificate Credibility	GIA、IGI、Other	Assessing Market Trust & Data Reliability Across Certification Systems

Within the DAII™ model, the traditional 4Cs represent an important part of the core quality evaluation, but not the entirety. Proportions and optical geometry factors account for approximately one-third of the overall model, demonstrating DAII's emphasis on the structural quality of a round diamond.

Layer 1: Foundational Quality Data

The first layer includes Color, Clarity, and Cut. These three items correspond to the most direct core data affecting quality evaluation in the traditional 4Cs. Color reflects color grade purity, Clarity reflects the impact of inclusions and surface characteristics on quality, and Cut provides a basic assessment of the cutting grade.

However, in DAII™, these three items are merely foundational inputs for the model rather than final conclusions. This is because Color, Clarity, and Cut all fall into graded ranges rather than precise numerical values. DAII converts these grades into standardized sub-scores, which are then evaluated alongside other structural data.

Layer 2: Cutting Execution

The second layer includes Polish and Symmetry. Polish reflects the finished state of the diamond's facet surfaces, while Symmetry reflects facet arrangement, geometric symmetry, and overall structural stability. Although these two items are often treated as supplementary data, they carry practical significance in round diamond quality evaluation.

Two diamonds both graded as Excellent Cut can still differ in facet precision, visual consistency, and light performance if one possesses more stable Polish and Symmetry while the other barely meets the grading boundary. Therefore, DAII™ incorporates cutting execution independently into the model instead of relying solely on the Cut grade.

Layer 3: Proportions and Optical Geometry Data

The third layer is where DAII™ connects most deeply with Tolkowsky's round diamond proportion philosophy.

A round diamond's Table, Depth, Crown Angle, Pavilion Angle, Crown Height, and Pavilion Height directly affect how light enters, reflects, refracts, and returns. This data is not secondary information; it is the core of the round diamond's optical structure.

In DAII™, proportions are viewed as a set of geometric relationships rather than isolated numbers. The Table affects the face-up visual proportions and the way light enters and exits; the Depth affects the face-up diameter and the efficiency of weight distribution; the Crown Angle and Crown Height affect fire and dispersion; and the Pavilion Angle and Pavilion Height directly affect light return and the risk of light leakage.

Proportion Parameter	High-Stability Range	Borderline Range
Table	Approx. 54.0–56.5%	Around 60.0%
Depth	Approx. 61.0–62.2%	Around 63.5%
Crown Angle	Approx. 34.2–34.8°	Around 36.0°
Pavilion Angle	Approx. 40.6–40.9°	Around 41.5°
Crown Height	Approx. 14.5–15.5%	Around 16.5%
Pavilion Height	Approx. 42.8–43.5%	Around 44.2%

In DAII™ model estimations, comparing two Excellent Cut diamonds—one with highly stable proportions and one within borderline ranges—can lead to a 10%–20% gap in their proportion sub-scores. This results in a 5%–15% difference in the final DAII™ Score.

This percentage does not indicate an exact visual difference in sparkle, which varies based on lighting and environment. Instead, it proves that diamonds with identical certificate grades can still possess distinct structural differences.

Layer 4: Structural and Visual Risk Factors

This layer includes Fluorescence, Girdle, Culet, and Ratio. While often overlooked in retail, these serve as essential risk-adjustment metrics in DAII™. For instance, an overly thin girdle impacts durability, an overly thick one reduces face-up size efficiency, and fluorescence can affect visual clarity. This layer prevents diamonds with high 4C grades but hidden structural risks from being over-evaluated.

Layer 5: Certificate Confidence

Grading strictness and market trust vary by gemological laboratory. DAII™ treats certificate confidence as a final reliability modifier rather than a simple data source, aligning the score with real-world market trust.

Ultimately, a high DAII™ Score requires balanced performance across all five layers, not just a single high grade. The 4Cs categorize a diamond; DAII maps its entire quality structure.

Score Interpretation: What the DAII™ Score Represents

The DAII™ Score is a data-driven model output that standardizes, weights, and adjusts multiple parameters alongside certificate confidence. It tracks a diamond's comprehensive quality rather than a single surface grade.

Because it evaluates the entire structure, a diamond with slightly lower 4C grades but superior geometric proportions can achieve a higher DAII™ Score than a poorly proportioned diamond with a higher surface grade.

Previously, separating two identical D / VS1 / Excellent diamonds required line-by-line certificate analysis and expert expertise. DAII translates these subtle variances into a consistent index. The score is not a pricing model or a direct buying recommendation; it simply answers:

Is this diamond structurally superior to another?
Does it hold a higher quality position within the same 4C grade?

Impact: How DAII™ Reshapes Diamond Comparison

DAII™ shifts diamond evaluation from single-grade linear comparisons (e.g., D > E) to multi-parameter structural analysis.

Real diamond quality is not linear; it relies on parameter combinations. A diamond with high Color but poor proportions may have lower overall quality than a diamond with slightly lower surface grades but superior geometric structure. DAII™ provides a unified framework to compare diamonds within a complete model, offering practical value for inventory management, quality sorting, and product screening.

Current Application: DAII™ and the 4Cs

DAII™ is initially applied to round diamonds because the round brilliant cut possesses the most mature, mathematically quantifiable proportional foundation (Table, Depth, Crown Angle, etc.).

DAII™ integrates the traditional 4Cs and certificate data rather than replacing them. The 4Cs and certificates serve as foundational inputs, moving diamond evaluation from static "grade reading" to dynamic "structural analysis."

Conclusion: A Modern AI Diamond Quality Index

DAII™ integrates foundational quality, proportional geometry, cutting execution, visual risks, and certificate confidence into a single quality index.

It advances diamond comparison from scattered data points to comprehensive structural analysis, allowing diamonds with identical 4C grades to display clear structural differences.

The core of DAII™ is not just to give a diamond another score, but to simply and comprehensively explain to consumers from multiple angles: why this diamond is more perfect!