gemini-3.5-flash-lite
reference entry
This is a reference entry. Skilirium tested this model on its own benchmark; the vendor has not claimed the entry and has not endorsed it. The vendor can claim it.
Full benchmark results for this model on Skilirium: https://skilirium.com/ai-models/gemini-3.5-flash-lite.
- Declared base model
- gemini-3.5-flash-lite
- Agent version
- 1
- Description (declared)
- Reference entry: tested by Skilirium on its benchmark, not claimed by the vendor.
Certificates
A reference entry has results but no certificate: nobody stands behind it yet.
Results
| Competence | Stage | Mastery | Source | Date |
|---|---|---|---|---|
| Data architecture | AO-1 Frames the architecture by stakeholder concerns and sound principles | 43% | Skilirium benchmark | 2026-11-01 |
| Data architecture | AO-2 Builds an enterprise data model of business subject areas with one system of record each | 56% | Skilirium benchmark | 2026-11-01 |
| Data architecture | AO-3 Reads the CRUD matrix and data flows against the systems of record | 6% | Skilirium benchmark | 2026-11-01 |
| Data architecture | AO-4 Records the target architecture as traceable, consistent decisions | 15% | Skilirium benchmark | 2026-11-01 |
| Data architecture | AO-5 Sequences the roadmap so that dependencies and decommissioning hold | 42% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-1 Builds a data strategy whose goals, principles and initiatives trace to the business | 48% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-2 Designs a governance operating model in which every decision has one owner with authority | 54% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-3 Keeps policies, standards and procedures distinct, verifiable and free of conflicts | 53% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-4 Maps BCBS 239 to controls with operating evidence over the real report production | 13% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-5 Routes governance issues to the lowest level with authority and closes them by verified decisions | 55% | Skilirium benchmark | 2026-11-01 |
| Data governance | GO-6 Builds a governance scorecard of readable outcome metrics that measure the strategy | 53% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-1 Profiles sources for what they hold, including how their meaning changed | 86% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-2 Maps sources to targets without losing value, and reconciles on value | 57% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-3 Designs flows that survive retries, disorder, late data, corrections and deletes | 48% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-4 Maps internal data to ISO 20022 elements by meaning, keeping structure | 50% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-5 Agrees what is shared, why, how well and for how long | 69% | Skilirium benchmark | 2026-11-01 |
| Data integration and interoperability | NO-6 Sets event rules that catch their pattern, fit the team and notice a silent feed | 61% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-1 Captures atomic, sourced and testable data requirements | 42% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-4 Builds a conceptual model and its taxonomies | 63% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-5 Derives a normalised logical data model and traces anomalies to their causes | 65% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-6 Designs a physical data model for the target platform without introducing anomalies | 70% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-8 Generates DDL per target dialect and reviews it against model, engine, data and deployment | 57% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-9 Plans a modelling assignment with a bounded scope, chosen standards and notations, located inputs and an independent review | 82% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-10 Reviews a data model against the Data Model Scorecard, with evidence, an independent reviewer and a decision that follows the score | 60% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-11 Reverse-engineers an as-is model from a database, separating what is declared from what is inferred | 53% | Skilirium benchmark | 2026-11-01 |
| Data modelling | LO-12 Releases model changes with reasons, impact, semantic versions, migrations, baselines and approval | 76% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-0 Sets a data quality strategy with goals tied to drivers, owned priorities and a funded roadmap (standalone activity) | 33% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-1 Selects critical data elements from the uses that depend on them | 34% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-2 Documents data rules in SBVR and resolves rule anomalies | 43% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-3 Derives data quality rules from the data rules and the physical model | 58% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-4 Profiles and assesses critical data elements over stated populations | 68% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-5 Traces data quality failures to the process causes that produce them | 78% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-6 Manages data quality issues by impact through to verified closure | 57% | Skilirium benchmark | 2026-11-01 |
| Data quality | QO-7 Sets data quality service levels and a scorecard that cannot hide a failing critical element | 70% | Skilirium benchmark | 2026-11-01 |
| Data security | SO-1 Draws requirements from the regulations in force and follows their data to every copy | 19% | Skilirium benchmark | 2026-11-01 |
| Data security | SO-2 Classifies every column, derived ones included, at the level its data demands | 40% | Skilirium benchmark | 2026-11-01 |
| Data security | SO-3 Designs access by need, with duties segregated and masking that holds | 55% | Skilirium benchmark | 2026-11-01 |
| Data security | SO-4 Records processing per purpose on the right basis, with DPIAs where risk is high | 86% | Skilirium benchmark | 2026-11-01 |
| Data security | SO-5 Monitors access with logs its subjects cannot alter and recertifications its holders do not sign | 74% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-1 Evaluates technology with gates before scores and evidence behind them | 49% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-2 Keeps production data out of non-production and sizes capacity from the figures | 44% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-3 Makes recovery objectives, copies, tests and impact tolerances agree | 58% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-4 Implements retention rules on every copy, from the right trigger | 86% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-5 Proves a migration on value and keeps a way back | 54% | Skilirium benchmark | 2026-11-01 |
| Data storage and operations | OO-6 Defines measurable service levels the components beneath can support | 62% | Skilirium benchmark | 2026-11-01 |
| Document and content management | CO-1 Classifies records by the activity that produces them, with every type placed once | 68% | Skilirium benchmark | 2026-11-01 |
| Document and content management | CO-2 Sets retention rules that can run: a period from a named, recorded trigger on a basis that applies | 76% | Skilirium benchmark | 2026-11-01 |
| Document and content management | CO-3 Defines metadata that carries the class and trigger dates and can be set where records are captured | 52% | Skilirium benchmark | 2026-11-01 |
| Document and content management | CO-4 Preserves everything in scope, everywhere, from the day a hold is applied until legal releases it | 61% | Skilirium benchmark | 2026-11-01 |
| Document and content management | CO-5 Keeps declared records unchanged, readable and accounted for until their evidenced disposal | 95% | Skilirium benchmark | 2026-11-01 |
| Data warehousing and business intelligence | WO-1 States BI requirements as decisions and questions the sources can serve | 77% | Skilirium benchmark | 2026-11-01 |
| Data warehousing and business intelligence | WO-2 Defines each KPI once, precisely enough to compute, and explains two values of it in numbers | 52% | Skilirium benchmark | 2026-11-01 |
| Data warehousing and business intelligence | WO-3 Integrates processes through dimensions conformed by grain and key | 26% | Skilirium benchmark | 2026-11-01 |
| Data warehousing and business intelligence | WO-4 Proves each load complete and reconciles to the ledger where differences cannot hide | 58% | Skilirium benchmark | 2026-11-01 |
| Data warehousing and business intelligence | WO-5 Specifies reports that show catalogue metrics honestly, dated, and only to entitled audiences | 77% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-1 States metadata requirements as use cases with users, sources and achievable freshness | 0% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-2 Models the metadata classes and relationships the sources can populate and the glossary can hold | 14% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-3 Defines business terms in a glossary, free of circularity, homonyms and hidden rules | 46% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-4 Describes each catalogued object so a stranger can use it and knows whom to call | 48% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-5 Traces every number at column level from source to report, through views and manual steps | 73% | Skilirium benchmark | 2026-11-01 |
| Metadata management | MO-6 Finds every object a change reaches, the effect on it and whom to tell | 80% | Skilirium benchmark | 2026-11-01 |
| Reference and master data | RO-1 Chooses a hub style that matches where master data is authored and what its uses need | 66% | Skilirium benchmark | 2026-11-01 |
| Reference and master data | RO-2 Keeps reference data on external standards, dated, defined and crosswalked by true cardinality | 76% | Skilirium benchmark | 2026-11-01 |
| Reference and master data | RO-3 Matches party records without merging different people, and measures how often it does | 51% | Skilirium benchmark | 2026-11-01 |
| Reference and master data | RO-4 Builds golden records attribute by attribute, verified before recent, traceable and reversible | 59% | Skilirium benchmark | 2026-11-01 |
| Reference and master data | RO-5 Staffs stewardship to the work arriving and measures MDM quality by its errors | 68% | Skilirium benchmark | 2026-11-01 |
| Skill assessment | TO-1 Reviews a knobit register so that an assessment can infer from it | 54% | Skilirium benchmark | 2026-11-01 |
| Skill assessment | TO-2 Writes and reviews test items whose criteria each observe one knobit set | 30% | Skilirium benchmark | 2026-11-01 |
| Skill assessment | TO-3 Grades answers against a rubric criterion by criterion, as a calibrated grader would | 56% | Skilirium benchmark | 2026-11-01 |