In today’s interconnected financial landscape, robust KYC (Know Your Customer) procedures are essential to prevent identity fraud, terrorist financing, and money laundering. As criminal networks evolve, so do their techniques for bypassing traditional controls — leveraging false identities, forged documents, and complex transaction layers to avoid detection.

To counter these threats, global regulators have strengthened KYC and Customer Due Diligence (CDD) requirements. Financial institutions are now mandated to verify customer identities, assess risk profiles, and monitor ongoing transactions as part of a holistic compliance strategy. In Malaysia, for instance, the Anti-Money Laundering, Anti-Terrorism Financing and Proceeds of Unlawful Activities Act (AMLA) enforces stringent obligations on reporting institutions. These include Enhanced Due Diligence (EDD) for high-risk customers and timely escalation of suspicious cases.

However, traditional KYC processes are often manual, slow, and error-prone — particularly when reviewing large volumes of documents or conducting repeated compliance checks. These inefficiencies lead to increased costs, delayed onboarding, and potential compliance risks.

To address these challenges, forward-thinking organizations are turning to Agentic Automation — a combination of RPA and intelligent agents that replicate human decision-making. By integrating technologies like UiPath’s Document Understanding, long-running workflows, and AI-driven agents, institutions can automate identity verification, cross-check applicant information with internal and external data sources, and apply contextual reasoning to flag suspicious behavior.

This demo showcases how a KYC review and approval process can be streamlined end-to-end using Agentic Automation in UiPath — reducing manual effort, improving consistency, and ensuring more accurate compliance decisions.

Know Your Customer (KYC) is a cornerstone of Anti-Money Laundering (AML) compliance, requiring financial institutions to verify customer identities, screen for risks, and assess the legitimacy of financial activities. Despite its importance, many organizations continue to rely on fragmented and manual processes to fulfill these obligations. From the moment a customer submits identity documents – such as an IC, utility bill, or bank statement – the process often requires human staff to open emails, download attachments, and visually inspect each file. Data must be manually extracted and cross-referenced, for example, comparing names and IC numbers across documents to ensure consistency. Compliance checks, such as verifying against internal blacklists or external PEP (Politically Exposed Person) and sanctions lists, are also handled separately, often through isolated systems that require multiple logins or manual API calls.

This disjointed approach leads to several challenges. Onboarding timelines stretch out, as even straightforward applications are slowed down by repetitive administrative work. Compliance teams are overwhelmed with low-value tasks, leaving less time to investigate truly suspicious behavior. Errors are common, whether from oversight, fatigue, or inconsistent application of business rules. As case volumes rise, institutions are forced to hire more analysts just to keep up with basic checks – driving up operational costs while increasing the risk of compliance breaches.

The financial institution in this case study was grappling with exactly these issues. Every new customer file triggered a lengthy manual workflow, requiring multiple steps of document handling, validation, screening, and in many cases, escalation to human reviewers. While the organization was committed to regulatory compliance, its legacy approach made it difficult to scale operations, adapt to new guidelines, or ensure consistent decision-making. The organization needed a smarter, faster way to handle the end-to-end KYC process – one that could reduce human workload, minimize delays, and improve the accuracy and defensibility of compliance decisions.

The primary objective for the financial institution was to transform its Know Your Customer (KYC) operations into a more intelligent, scalable, and resilient process. As regulatory expectations continue to tighten and financial crime tactics evolve, the institution recognized that traditional, manual approaches to KYC verification were no longer sustainable. They sought to automate the end-to-end KYC lifecycle – from document collection and validation to compliance screening and decision-making – using a combination of Robotic Process Automation and intelligent agents.

The institution’s vision was to reduce dependency on human intervention for routine verification tasks, enhance the consistency and speed of compliance checks, and create a framework that could adapt dynamically to changing regulatory policies or internal risk thresholds.

Key objectives and goals included:

1. Automate the entire KYC workflow: Enable straight-through processing by using RPA to handle email intake, document extraction, rule-based validation, API checks, and communication with human agents.
2. Improve decision accuracy: Leverage AI agents and validation models to reduce human error, flag inconsistencies in documentation, and identify high-risk customers more effectively.
3. Streamline compliance checks: Integrate internal PEP screening API, along with internal business rules, to ensure timely and consistent due diligence across all applications.
4. Reduce manual workload: Eliminate repetitive, time-consuming tasks so compliance analysts can focus on escalated or complex cases that require human judgment.
5. Ensure full regulatory compliance: Maintain auditability and traceability across every step of the KYC process, aligning with AMLA, FATF, and local regulatory standards.
6. Enable scalable operations: Deploy a modular, low-code automation solution that can easily accommodate increased onboarding volume, rule changes, or new risk parameters without major rework.

To tackle the financial institution’s AML challenges, Cybiant designed a cutting-edge Agentic Automation solution leveraging the UiPath RPA platform combined with advanced AI capabilities. Central to the solution is an intelligent agent powered by a standard OpenAI Large Language Model (LLM). This model was further trained and contextually grounded on the specific AML regulations of Bank Negara Malaysia, ensuring the agent’s recommendations and flagging are fully compliant and tailored to local regulatory standards.

Cybiant developed a scalable ingestion system capable of processing high volume of daily transactions from the client’s transaction statements. This data pipeline enables the intelligent agent to analyze transaction flows continuously, identifying suspicious patterns and anomalies that may indicate potential money laundering activities. By embedding the agent within the existing UiPath RPA workflows, Cybiant enabled a seamless, automated end-to-end process—significantly reducing manual review efforts and accelerating compliance workflows.

To complement this, Cybiant also implemented rule-based logic within UiPath Studio to automate traditional compliance checks such as name matching, document validity verification, and identification number formatting. These rules are fully configurable, allowing the institution to adjust the compliance strictness depending on the customer risk level, while maintaining auditability and alignment with regulatory expectations.

Key steps Cybiant took to create this solution included:

1. Risk Criteria Definition and Model Contextualization: Defined key compliance parameters by incorporating a list of high-risk countries and Malaysia’s official sanctions list. These elements were embedded into the agent’s prompt to provide contextual awareness of geopolitical and sanction-related risks. The model’s responses were shaped through prompt engineering to align with internal compliance policies and risk indicators relevant to Malaysia’s regulatory environment.
2. Agent Development: Built an intelligent agent on the UiPath platform that leverages the trained LLM to assess transactions and flag suspicious activity using globally recognized patterns defined by AMLA and FATF.
3. Data Ingestion Pipeline: Engineered a robust ingestion system capable of processing and normalizing high volume daily transactions from diverse financial data sources, allowing real-time transaction monitoring.
4. Rule-Based Compliance Checks: Developed and deployed rule-based validation workflows in UiPath Studio to handle identity verification, document matching, and static compliance rules such as expiry dates and name similarity thresholds.
5. Integration with RPA Workflows: Embedded both the intelligent agent and rule-based checks into the client’s existing UiPath workflows to automate end-to-end KYC and AML compliance processes, including document screening, case creation, and escalation.
6. Testing and Validation: Conducted extensive testing to validate the solution’s detection accuracy, reduce false positives, and confirm that it met regulatory reporting and audit trail standards.
7. Deployment and Monitoring: Successfully deployed the solution into a live production environment with ongoing monitoring and adaptive tuning, ensuring continued alignment with evolving financial crime tactics and compliance expectations.

Cybiant deployed a strategic combination of Robotic Process Automation (RPA) and intelligent agent technology to automate and enhance the Know Your Customer (KYC) and Anti-Money Laundering (AML) compliance process. The core automation was developed using UiPath, which orchestrates tasks such as document validation, name matching, expiry checks, and blacklist screenings. All of these validations are handled using rule-based logic in UiPath Studio, ensuring precision, traceability, and strict adherence to regulatory rules.

In parallel, Cybiant integrated an intelligent compliance monitoring agent powered by a standard OpenAI Large Language Model (LLM). The model was guided by a clearly defined prompt designed for the financial institution’s compliance context. This prompt reflects regulatory expectations under Malaysia’s AMLA, FATF guidelines, and internal policy. It instructs the agent to analyze customer profiles and transaction data for suspicious behaviors—such as structuring, rapid fund movements, inconsistencies based on occupation, or links to high-risk countries.

The agent evaluates incoming data sets including transaction history, customer metadata, and risk profiles. It cross-references with internal risk lists (e.g., Malaysia’s sanctions list and high-risk countries), and generates concise risk summaries such as: “Structuring behavior detected: multiple RM10,000 transactions within 3 days” or “Customer appears in Malaysia sanctions list.” If no issues are found, the agent returns “No suspicious transactions detected.”

This AI-powered assessment is embedded within the broader UiPath workflow, enabling a smart, efficient, and auditable process. It significantly reduces manual review time while enhancing risk visibility. Together, the UiPath automation and OpenAI agent form a cohesive compliance automation system that aligns with both local regulatory standards and global best practices.

Cybiant’s Agentic Automation solution revolutionized the institution’s Know Your Customer (KYC) and transaction monitoring processes. By combining intelligent agents, prompt-based risk assessment, and robust RPA workflows in UiPath, the institution achieved faster compliance checks, higher accuracy, and greater operational agility—without the overhead of complex model training or manual reviews.

Key results and benefits included:

1. Streamlined Transaction Monitoring: Automated the screening of transaction histories using agentic logic, enabling rapid identification of suspicious patterns with minimal human intervention.
2. Faster Compliance Review: Reduced processing time per customer profile by automating repetitive KYC tasks and flag evaluation, accelerating the overall due diligence process.
3. Improved Accuracy and Consistency: Leveraged prompt-engineered AI to apply consistent risk checks, reducing subjectivity and human error in identifying high-risk transactions.
4. Integrated Sanctions Screening: Embedded Malaysia’s sanctions list directly into the workflow, ensuring every customer and transaction was automatically cross-checked against critical regulatory watchlists.
5. Flexible Risk Configuration: Allowed for easy adjustment of thresholds (e.g., transaction limits, name-matching logic) directly in UiPath, giving compliance officers control to adapt policies based on evolving risk appetites.
6. Reduced Operational Load: Minimized manual review and rule management effort by centralizing compliance logic into a digital agent framework.

Overall, Cybiant’s solution empowered the institution to run smarter and more adaptive KYC processes, improving both compliance outcomes and team efficiency. The use of practical AI – guided by real-world policy inputs rather than complex model retraining – ensured a scalable and audit-ready solution aligned with Malaysia’s regulatory environment.