OUR SERVICES
Three Services, Clearly Scoped
Each of Elden's services addresses a specific AI application type. This page explains what each one involves, what is included at delivery, and who it is best suited for.
Back to HomeOUR METHODOLOGY
How We Approach Each Engagement
Every project follows a consistent structure: scoping, data review, build, testing, and delivery. The specifics vary by service type, but the underlying discipline is the same — understand the problem before building a solution.
We keep our process transparent throughout. You will know what stage we are at, what the current outputs look like, and what decisions still need to be made.
Quality checks are built into the timeline, not added at the end. This means the final delivery is something both sides have already reviewed — not a surprise package.
Scoping Call
Understand your data, goals, and constraints
Engagement Document
Written scope agreed before work begins
Build & Iterate
Regular check-ins during delivery
Delivery & Handover
Documentation, walkthrough, support window
AI for Claims Processing
Development of automated systems for insurance or warranty claims — covering document intake, information extraction, initial assessment, and routing to appropriate review teams. The service includes decision-tree modelling, confidence scoring, and human review workflow integration. Designed for insurers or manufacturers handling high claim volumes.
WHAT IS INCLUDED
- Document intake pipeline configuration
- Information extraction model (trained on your claim types)
- Decision-tree and confidence scoring logic
- Human review routing workflow
- Documentation and walkthrough session
PROCESS STEPS
- W1–2Data review and field mapping
- W3–6Extraction model training and validation
- W7–8Decision logic and routing build
- W9–10Testing, documentation, and handover
Topic Modelling Service
Building models that automatically discover and track themes across large collections of documents, articles, or conversations. The service covers preprocessing, model selection, topic labelling, and temporal trend tracking. Suited for research organisations, media companies, or customer insight teams.
WHAT IS INCLUDED
- Text preprocessing pipeline
- Model selection and evaluation (LDA, NMF, or embedding-based)
- Topic labelling and interpretation
- Temporal trend analysis output
- Documentation and walkthrough session
PROCESS STEPS
- W1Corpus review and preprocessing design
- W2–3Model training and comparison
- W4–5Topic labelling and trend tracking
- W6Output delivery and documentation
AI Opportunity Mapping
A discovery engagement that systematically identifies where AI can add value across your organisation's functions. The process includes departmental interviews, process analysis, and opportunity scoring against effort and impact. Produces a visual opportunity map with prioritised initiatives.
WHAT IS INCLUDED
- Structured departmental interviews
- Process and data landscape review
- Effort vs. impact scoring matrix
- Visual opportunity map (PDF and editable format)
- Written summary with prioritised initiatives
PROCESS STEPS
- W1Stakeholder interviews and process review
- W2Opportunity analysis and scoring
- W3Map preparation, written summary, and presentation
CHOOSING THE RIGHT SERVICE
Which Service Fits Your Situation?
| CLAIMS PROCESSING | TOPIC MODELLING | AI MAPPING | |
|---|---|---|---|
| Best for | Insurers, warranty providers | Research, media, insight teams | Any organisation exploring AI |
| Duration | 10 weeks | 6 weeks | 3 weeks |
| Price (RM) | 4,000 | 2,300 | 850 |
| Requires existing labelled data | |||
| Suitable as first AI engagement | Sometimes | ||
| Delivers a deployed system |
Not sure which applies to you? The AI Opportunity Mapping engagement is the natural starting point.
ACROSS ALL SERVICES
Technical & Professional Standards
Data Security
ISO 27001-aligned data handling. Access limited to project requirements.
Performance Testing
All models are evaluated against pre-agreed performance criteria before delivery.
Full Documentation
Every delivery includes technical documentation suitable for your internal team to maintain.
Post-Delivery Support
A defined support window follows every project for questions and minor fixes.
Ethical Review
Bias and fairness checks are conducted on models where decisions affect individuals.
Reproducible Builds
Model training is logged and version-controlled so results can be reproduced and audited.
Have Questions About a Specific Service?
We are happy to discuss whether any of our services would suit your situation before you commit to anything. A scoping call is free and there is no obligation.
Start a Conversation