Use Cases

SPICE solves real-world challenges

Across infrastructure providers and market compliances, SPICE helps teams reduce the false positives and make faster, governed, agentic decisions they can stand behind.

Book a demoExplore the platform
Compliance teams

Compliance Workflow Platform

The problem

Compliance teams need to monitor large volumes of unstructured information and identify issues that could affect their regulatory obligations. Documents arrive continuously across multiple systems and reviewing them manually is slow and difficult to scale. Teams also need to demonstrate what was reviewed, what was identified and how each issue was handled.

What SPICE does

SPICE provides a secure compliance workflow platform for industries that rely on unstructured data. It ingests PDFs and other documents in real time by connecting to the systems where information becomes available. Agents extract and structure the relevant information before applying compliance rules and raising findings for review.

SPICE manages the process from document ingestion through to review and resolution. Every finding retains its source information and decision history. Compliance teams remain in control through human review and workflow-based approval.

  • Real-time ingestion of PDFs and other unstructured documents
  • Automated extraction and structuring of compliance information
  • Compliance workflows that route findings for human review
  • Full provenance from source document to decision
  • Secure management of compliance data and evidence
  • Complete audit trail for every finding and action
  • Exportable evidence to support compliance obligations
Show case study
Bursa Malaysia

The challenge

The Malaysian stock exchange processes thousands of company announcements and regulatory filings on a regular basis. These documents contain critical information that compliance teams need to monitor. Reviewing large volumes of unstructured documents manually is resource intensive, particularly when compliance teams need to respond quickly.

How SPICE is used

SPICE connects directly to the stock exchange's source of company announcements and ingests new documents as they become available. Agents extract relevant information from each document and convert it into structured data.

This structured data is then used to power compliance alerts, support automated investigation and provide analytics across the full body of company announcements. Compliance teams can quickly identify issues that require attention and investigate them without manually searching through large volumes of documents.

Findings are routed through controlled workflows for review and resolution. Every finding retains its connection to the original announcement and the information used to generate it. This provides a clear chain of provenance from the source document through to the compliance action.

What was achieved

  • Manual compliance review reduced by approximately 90% from around ten days to one
  • Compliance alerts generated automatically from newly available announcements
  • Automated investigation supported by structured information extracted from source documents
  • Analytics enabled across the full body of company announcements
  • Complete provenance from source announcement through to review and resolution
  • Controlled workflows that keep human reviewers in the process
  • Every finding traceable to the source information used to generate it
~90%Less manual effort
~1 dayReview cycle, from ~10
FullAudit trail
Compliance alerts generated from company announcements
Surveillance

Agentic AI

The problem

Organisations hold valuable information across large databases and other data sources. Finding answers often requires specialist knowledge of the underlying systems and data structures. This makes complex enquiries and investigations slow and difficult for operational teams to perform themselves.

What SPICE does

SPICE provides a governed platform for deploying AI agents that can work directly with an organisation's data. Users can ask questions or describe a task in plain language and the agent determines how to find the answer or complete the work.

Agents can query and traverse complex databases and combine information from multiple sources. They can investigate anomalies and answer ad hoc questions or carry out repeatable tasks. Domain knowledge and system knowledge are captured as reusable skills so agents can work effectively within a specific operational environment.

The same platform can support very different use cases. This includes helping maintenance teams investigate rail assets and helping telecommunications teams understand transmission tower usage and anomalies.

  • Conversational access to complex organisational data
  • Automated database querying and investigation
  • Investigation of complex relationships and anomalies
  • Reusable skills that capture domain and system knowledge
  • Support for ad hoc enquiries and repeatable tasks
  • Governed agent execution within existing security controls
  • Human oversight and control throughout the workflow
  • Extensible across different industries and operational environments
Show case study
eflow

The challenge

The eflow PATH platform supports trading firms and other financial institutions in meeting their trade compliance obligations. Compliance analysts need to investigate large volumes of trading activity and alerts. Investigations can require analysis across large and complex datasets which makes it difficult for analysts to quickly establish what happened.

This creates a significant manual workload. Analysts can spend substantial time investigating alerts that ultimately prove to be false positives.

How SPICE is used

DHI provides eflow with an agentic capability that allows compliance analysts to investigate trading activity using natural language. Analysts can ask questions about trading behaviour and the agent investigates the underlying data to find the relevant information.

The agent can query and traverse complex trading databases to investigate patterns and relationships that may not be immediately apparent. It can also investigate individual trading alerts and help analysts determine whether the activity warrants further attention.

This gives analysts a flexible way to perform ad hoc investigations without manually constructing database queries. The same capability can also support repeatable investigation workflows and other compliance tasks.

The capability saves compliance analysts significant time by either supporting investigation through ad hoc database queries or by helping analysts to rapidly deprioritise false positives.

What was achieved

  • Complex trading investigations performed using natural language
  • Analysts able to interrogate large trading datasets without manually constructing queries
  • Faster investigation of trading alerts
  • Rapid deprioritisation of false positives
  • Significant reduction in manual investigation effort
  • Reusable agentic capability supporting multiple compliance workflows
SPICE agent investigating trading activity in natural language
Critical-infrastructure operators

AI Automation

The problem

Many operational processes require people to gather information from multiple sources and apply judgement before an outcome can be produced. These tasks can be difficult to scale when they involve large datasets, qualitative information or uncertain evidence.

What SPICE does

SPICE combines structured data, database agents and qualitative information to automate complex operational tasks. Governed agents can investigate information, apply defined rules and produce outcomes while keeping people in control of the process.

This enables organisations to automate repeatable tasks such as assessing risk, generating asset health indexes or processing low-risk exceptions. AI can handle the scale and complexity of the underlying information while rules and human oversight ensure that automated outcomes remain controlled and defensible.

  • Structured and unstructured information combined in a single process
  • Database agents that investigate and analyse complex datasets
  • Qualitative and uncertain information incorporated into assessments
  • Rules-based AI that operates within defined boundaries
  • Human-in-the-loop workflows for governed automation
  • Repeatable automation across large volumes of data
  • Evidence and reasoning retained behind automated outcomes
  • Outcomes that can trigger further workflows or actions
Show case study
DT Infrastructure

The challenge

DT Infrastructure holds vast amounts of structured data relating to its rail assets and their operational history. The challenge was turning this data into information that could support better asset management and maintenance decisions across a large network.

Analysing individual assets in detail can require significant human effort. Applying the same level of analysis across a large asset base would be impractical using manual processes alone.

How SPICE was used

SPICE brought together the available asset information to create a foundation for analytics and more informed decision making. The data could be analysed across individual assets as well as across areas and groups of related assets.

This also unlocked the ability to use AI agents to investigate the underlying asset data at scale. Agents can traverse the asset database and examine the history and characteristics of individual assets to support detailed health assessments.

The approach provides a pathway to generating detailed Asset Health Index information across a large asset base. This would otherwise require an impractical amount of manual analysis.

The same data foundation can also support the identification of patterns across assets and provide maintenance teams with a more targeted view of where further investigation or intervention may be required.

What was achieved

  • Vast quantities of structured asset data brought together for analysis
  • Analytics enabled across individual assets and the wider network
  • More informed asset management and maintenance decisions
  • AI agents able to investigate complex asset information at scale
  • Foundation established for detailed Asset Health Indexing
  • Asset analysis that would have been impractical through manual effort alone
1000+Assets analysed on a single platform
Better history tracking for asset forecasting linkage
SPICE asset health index generated from asset data
Get started

Find the use case for your mandate

Tell us the decision your team owns. We'll configure a SPICE demo around it, governed, explainable and ready for scrutiny.

Book a demoSee the platform