Deterministic Evaluation
A rule-driven evaluation engine applies explicit policy and quality rules instead of relying on holistic AI judgment alone.
Flagship project · Applied AI & ServiceNow
SNOWLens-AI: turning ServiceNow ticket data into evidence-backed operational intelligence.
SNOWLens-AI is a ticket intelligence platform designed to analyze ServiceNow support data using deterministic policy evaluation, evidence-grounded AI reasoning, quality scoring, and role-specific operational dashboards.
The problem
Ticket quality is difficult to measure consistently at scale.
Service teams generate large volumes of ticket data, but turning that data into meaningful quality and operational insight requires more than simply searching for keywords or asking an AI model for a score.
Policies need to be translated into measurable rules, evidence needs to be traceable, and AI-generated conclusions need to remain grounded in what actually happened on the ticket.
The approach
Separate what can be determined from what needs AI reasoning.
SNOWLens-AI combines deterministic evaluation with AI-assisted analysis rather than treating the entire ticket as an unrestricted AI reasoning problem.
Evidence-Grounded AI
AI narrative and review generation are grounded in ticket facts, activity history, policy evidence, and validated evidence references.
Policy & Quality Framework
Policies are normalized into structured rules that can be evaluated consistently across ServiceNow tickets.
Operational Dashboards
Director, Team Lead, and Engineer views provide different levels of insight into ticket quality, ownership, and operational performance.
Evaluation pipeline
From raw ticket data to validated intelligence.
ServiceNow Data
Ticket exports and operational data provide the factual foundation for analysis.
Evidence & Normalization
Ticket facts, activity history, policy data, and evidence are normalized into structured models.
Rule-Driven Evaluation
Deterministic policy and quality rules evaluate measurable compliance before AI reasoning is applied.
AI Reasoning
AI generates narrative and review content using the structured evidence and deterministic evaluation results.
Validated Results
AI outputs are validated for evidence references and scoring integrity before being presented.
Engineering
Built as a real application, not an AI demo.
The platform includes application infrastructure, persistence, authentication, administration, evaluation services, AI provider abstraction, and a full frontend experience.
✓
ASP.NET Core / .NET backend
✓
React + TypeScript frontend
✓
SQL Server persistence
✓
Entity Framework Core
✓
Role-based authentication and administration
✓
AI provider abstraction
✓
Deterministic rule evaluation architecture
✓
Evidence-grounded AI generation
✓
Automated backend and frontend testing
What it enables
A more explainable approach to ticket quality intelligence.
✓
Move ticket quality analysis from subjective review toward repeatable evaluation
✓
Separate deterministic policy compliance from AI-assisted reasoning
✓
Give leadership actionable visibility into ticket quality and operational patterns
✓
Provide engineers with evidence-backed feedback instead of opaque AI scores
✓
Create an architecture that can evolve across multiple AI providers
✓
Keep AI reasoning grounded in structured ticket evidence
Flagship project
SNOWLens-AI is the foundation. SNOWLens-Orchestrator extends the platform into operational processing.
The two projects are designed to work together: SNOWLens-AI provides the intelligence and evaluation layer, while SNOWLens-Orchestrator focuses on repeatable batch processing and operational control.