Scripted. Limited. Siloed.
- Pre-defined decision trees
- No real banking system access
- Single-turn interactions
- No governance or audit trail
NOOR — Agentic AI Banking Engagement Platform
Deploy customer and employee AI agents, transaction intelligence, knowledge automation, and governed banking workflows inside your bank-controlled environment.
Not A Chatbot
NOOR Isn't A Chatbot. It's An Agentic Banking AI.
NOOR is not a chatbot. It is an agentic AI banking engagement platform designed to understand customer context, retrieve bank-approved knowledge, analyze transactions, assist employees, support relationship managers, and orchestrate governed banking workflows through integration with the bank's systems.
Context-Aware
Action-Capable
Governed
Bank-Integrated
Contextual. Agentic. Governed.
- Multi-agent orchestration engine
- Live core banking API access
- Multi-turn memory & context
- Full audit trail & PII masking
Key Capabilities
Built For Bank-Grade AI
Bank-Controlled Deployment
NOOR is designed to run within the bank-controlled environment, supporting on-premise, private cloud, or hybrid deployment models based on bank requirements.
Data Residency
Customer data, prompts, embeddings, logs, and sensitive banking information remain under the bank's control according to the selected deployment architecture.
Arabic-First, Multilingual
NOOR is designed for Arabic and English banking experiences, with multilingual expansion supported based on bank requirements.
Core-Agnostic Integration
Through UBIL and bank APIs, NOOR connects to core banking, digital banking, CRM, cards, payments, onboarding, contact center, and enterprise knowledge systems.
LLM-Agnostic Architecture
NOOR can be configured with compatible model options selected according to the bank's infrastructure, governance, and security requirements.
Governed AI Operations
NOOR supports policy controls, access management, auditability, monitoring, and bank-defined approval flows for sensitive actions.
AI Agent Suite
Customer-Facing And Employee-Facing Agents
Customer Agents
Executes governed banking actions on the customer's behalf — transfers, card controls, standing orders and limit changes — each one routed through the bank's own approval and audit flow.
Answers product, fee, branch and policy questions using only bank-approved knowledge sources, with citations back to the underlying document.
Analyses spending, income and cash-flow patterns to surface pace alerts, overspend risk, subscription duplication and projected savings.
Matches merchant offers, loyalty rewards and travel deals to the customer's actual transaction profile rather than a broadcast campaign list.
Guides the customer through eligibility, required documents and application steps — reducing drop-off and inbound call volume.
Employee Agents
Prepares portfolio context, next-best-conversation prompts and product fit for the RM before a client meeting, drawn from CRM and core data.
Gives the advisor a live summary of the customer's position, the likely reason for the call and the approved resolution path — cutting handle time.
Keeps product, pricing and eligibility rules consistent across every channel, so front-line staff quote from the same governed source.
Retrieves internal policy, procedure and compliance guidance in seconds, with role-based access so staff only see what they are cleared to see.
How NOOR Works · Architecture
From Interaction To Governed Action
NOOR connects channels, agents, knowledge, core systems, and governance into a single bank-controlled loop — select any layer to explore it.
Channels
Layer 1 of 6Customer or employee initiates an interaction through any bank-connected channel.
Agent Orchestration
Layer 2 of 6Intent is classified and routed to the right specialised agent, which plans a multi-step response and holds turn-by-turn memory of the conversation.
Knowledge & Vector Store
Layer 3 of 6Bank-approved policy, product and procedure documents are embedded and retrieved with citations, so every answer traces back to an internal source.
UBIL & Bank APIs
Layer 4 of 6The integration layer normalises vendor payloads into one canonical banking model, shaping and authorising traffic before it reaches a system of record.
LLM Inference
Layer 5 of 6Model selection is the bank's decision. NOOR runs with compatible model options chosen against the bank's infrastructure, governance and security requirements.
Governance & Monitoring
Layer 6 of 6Prompt, response, workflow and action logs feed monitoring dashboards for usage, errors, escalations, response quality and risk flags.
Use Cases By Buyer
Tailored For Every Banking Stakeholder
Salary Delay Insight
NOOR detects that a customer's salary has not arrived on its usual date and explains the potential cash-flow impact before upcoming payments are due.
Subscription Audit
NOOR identifies recurring payments and helps the customer understand where money is being spent every month.
Loan Application Guidance
NOOR guides the customer through eligibility, required documents, and application steps — reducing drop-off and call volume.
RM Client Brief
The relationship manager opens a client file and receives portfolio context, recent activity and next-best-conversation prompts drawn from CRM and core data.
Internal Knowledge Lookup
Staff retrieve procedure and compliance guidance in seconds, with role-based access so each employee only sees what they are cleared to see.
Product & Pricing Answers
Product, fee and eligibility rules are quoted from one governed source, so every channel gives the customer the same answer.
Live Call Summary
The advisor sees the customer's position, the likely reason for the call and the approved resolution path before the conversation begins.
Self-Serve Resolution
Routine balance, statement and card queries resolve in the digital channel, so the queue keeps the calls that genuinely need a human.
Escalation Signals
Monitoring flags low-confidence answers and repeated escalations, giving the contact-center team a clear improvement backlog.
Credible AI Strategy
A defensible position on AI that avoids vendor dependency and does not bet the bank on a multi-year core replacement.
Adoption & ROI Reporting
Usage, deflection, response accuracy and escalation rates reported against the KPIs agreed at pilot.
Time-To-Market
Intelligence ships on the estate the bank already runs, so differentiation arrives in months rather than programme cycles.
Workflow Automation
Governed banking workflows execute end to end, with approval steps inserted wherever the bank's policy requires a human decision.
Environment Separation
DEV, SIT, UAT, PROD and DR stay separated and evidenced, so operations can promote changes with a clear audit position.
Monitoring Dashboards
Usage, errors, escalations, response quality and risk flags are visible in one place instead of stitched together after the fact.
Governance And Controls
Controls The Risk Team Can Sign
Access Control
- Role-based access for employees, administrators, and specialized agent roles.
- Approval workflows for sensitive operations, financial transactions, and regulated actions.
Compliance
- Prompt, response, workflow, and action audit logs for review and compliance.
- Policy guardrails that define what each agent can and cannot do.
Security
- PII handling aligned with bank data governance requirements.
- Environment separation across DEV, SIT, UAT, PROD, and DR where required.
Operations
- Model and prompt version control for controlled AI operations.
- Monitoring dashboards for usage, errors, escalations, response quality, and risk flags.
Deployment Model
From Pilot To Production
A structured, phased approach to onboarding agentic AI inside your secure, bank-controlled infrastructure. We partner with you at every stage.
01
Pilot
Controlled rollout with select teams, quick time-to-value, and measurable KPIs from day one to prove core platform value.
02
Enterprise Implementation
Full-scale deployment across business units, complete change management support, and phased, risk-mitigated rollout plans.
03
Environments
DEV, SIT, UAT, PROD, and DR environments fully separated, sandboxed, and managed to secure software development lifecycles.
04
Infrastructure
Flexible cloud-native or secure on-premise deployment with a highly scalable, fault-tolerant banking architecture.
05
Integration
Seamless, secure connection with core banking systems, digital frontends, CRMs, data lakes, and legacy enterprise workflows.
06
Support
Dedicated 24/7 banking-grade support, strict SLA-backed response times, proactive monitoring, and instant incident management.
Ready To Transform Your Banking?
Unlock high-fidelity AI automation and compliant engagement layers with NOOR, built specifically for regulated financial institutions.
Request A Demo