Unplanned interruptions
Plant operators absorb lost production time, urgent labor and parts costs, and schedule disruption when equipment fails unexpectedly.
Industrial reliability, made clearer
AI-assisted maintenance planning for industrial teams.
Unplanned equipment stoppages cost plants production time, urgent labor and parts spend. Reliability Desk is being developed to help maintenance and reliability teams organize work history, spot recurring signals and plan human-reviewed follow-ups.
Local-first product prototype available · production AI in development
Asset watch / concept previewIllustrative UI
Pump train P-04ReviewWork orders + inspection notes
Compressor C-12MonitorCondition signals + service history
Example interface only. No live equipment data is shown.
Built around the realities of industrial maintenance in Nigeria.
Plant and facilities teams
Maintenance and reliability leads
Asset-intensive operators
Try Reliability Desk
Work with the prototype, right here.
Add equipment, record maintenance events, review a transparent priority score, import CSVs and back up your local workspace.
Seed records are fictional; records you enter remain in your browser. Scores use an illustrative rule, not a live AI model or validated equipment-risk assessment.
The operating problem
When the warning is buried, downtime gets expensive.
Maintenance teams often have the clues, but they are spread across logs, inspections, spreadsheets and people’s memory.
Fragmented maintenance records
Work orders and inspection notes are difficult to compare when they live in disconnected formats and systems.
Reactive prioritization
Teams need a practical way to decide which asset deserves attention first, without replacing the judgment of experienced technicians.
The product direction
From scattered records to a focused maintenance queue.
The available demo shows how an asset queue, example maintenance signals and human-reviewed follow-ups could work together. A production product with customer data, validated AI and operational integrations remains in development.
Explore Reliability Desk+Spend less time searching fragmented work history
+Surface recurring faults before planning the next check
+Focus review on assets with stronger evidence signals
+Keep qualified technicians in control of decisions
Product capabilities
A practical loop for reliability teams.
The demo already supports the sample workflow below. Production data connections, validated AI analysis and live operational use are not yet available.
Maintenance record intake
Structure work orders, technician notes and inspection documents for consistent asset histories.
Recurring fault signals
Identify repeated failure descriptions and changes in condition indicators for human review.
Risk-aware work queue
Combine asset context, recurrence and available operating signals into explainable review priorities.
Inspection recommendations
Suggest follow-up checks and planning prompts that a qualified maintenance lead can approve.
Feedback from the field
Capture whether recommendations were useful and improve the workflow with pilot-team feedback.
Readable asset summaries
Generate concise summaries of recent work, recurring issues and open questions for shift handovers.
AI technology
Useful signals, not a black box.
AI is intended to help maintenance teams find patterns across records and telemetry. Model choices and performance targets will be validated with pilot data before any operational claims are made.
Data inputs
Planned inputs include work orders, inspection notes, asset metadata and condition or operating time-series where customers can provide them.
What it may detect
Repeated fault descriptions, unusual changes in sensor patterns and assets whose maintenance history warrants review.
Methods under evaluation
Natural-language processing for maintenance text, anomaly detection and time-series analysis. Specific model architectures remain subject to data quality and pilot validation.
Human oversight
Recommendations are intended as decision support. Technicians and reliability leads remain responsible for inspections, work authorization and safety decisions.
How it is intended to work
Start with the records teams already have.
Connect a pilot dataset
Agree on a narrow asset group and review available records, access permissions and data quality.
Build a usable asset history
Normalize equipment identifiers, service events and inspection findings into a common timeline.
Surface patterns for review
Apply validated analysis to highlight repeat issues and unusual signals with supporting evidence.
Measure and refine
Review outputs with maintenance staff and track agreed pilot measures before considering expansion.
Technology & infrastructure
Cloud and accelerated computing, applied where they help.
Infrastructure choices are planned, not yet deployed. The pilot architecture will be sized around data sensitivity, latency, workload and customer requirements.
Planned AWS and Microsoft Azure evaluation
AWS and Microsoft Azure are candidate platforms for encrypted file storage, managed asset records, API hosting, identity controls, monitoring and AI inference. On AWS, S3, RDS, ECS or Lambda, CloudFront, Bedrock and CloudWatch are options. On Azure, Blob Storage, Azure Database for PostgreSQL, Container Apps or Functions, Entra ID, Azure Monitor and Azure AI Foundry are options. The product team would select a primary architecture based on pilot security, data-residency, integration and workload requirements rather than assume a multi-cloud deployment.
Planned NVIDIA evaluation
GPU acceleration may support experimentation on larger maintenance datasets and faster model inference if pilot workloads justify it. CUDA, TensorRT and Triton Inference Server are candidates for optimizing validated models; smaller workloads may run efficiently without dedicated GPUs.
Market opportunity
For teams responsible for uptime and equipment health.
The initial customer focus is industrial plants, oil and gas operators, and facilities teams in Nigeria that manage recurring maintenance work. Expansion to other markets would follow validated pilots, local partnerships and operational readiness. No market-size figures are published until independently verified.
Primary usersMaintenance managers, reliability engineers and plant operations leads
Initial marketNigeria; other African markets are a later possibility
Revenue modelPlanned site subscription with optional onboarding and usage-based AI processing
Go-to-marketDiscovery conversations followed by scoped operator pilots
Current product status
Local-first prototype available
Users can create assets, log maintenance events, import CSV data, and export a portable workspace backup. Data stays in the current browser. Production AI, shared customer workspaces, verified pilots and deployments are not yet available.
Pilot measures to establish: time to find asset context, repeat-fault review coverage, useful recommendation rate and false-alert rate. Targets will be agreed after baseline measurement with pilot teams.
Why now
Industrial teams need to make more of the data they already collect.
Digital maintenance records and connected equipment are producing more operational information, but useful decisions still depend on bringing the pieces together. A carefully validated AI workflow can help teams focus attention without removing people from the loop.
Roadmap
Validate the workflow before scaling the technology.
Operator discovery
Confirm the highest-cost maintenance workflow and data access constraints.
MVP workflow
Prototype asset histories, maintenance search and a review queue.
Scoped pilot
Test analysis with consenting operators and human-reviewed outputs.
Measured deployment
Harden security and scale cloud or GPU workloads only after validation.
The people behind the work
Engineering experience, close to the operating floor.
Company leadership and operational roles listed from current company information. Software and data-science team credentials will be published when confirmed.
Founder / Chief Executive Officer
Gbenga Fasasi
Company leadership, strategic direction and business development at Queen Susa Limited.
Managing Director
Joseph Fasasi
Oversees day-to-day operations, project delivery and coordination of technical teams.
Technical operations
Field technical team
Site supervisors and skilled trades support maintenance, installation, fabrication and construction work.
Product development
Building the product team
The software engineering and AI development function is being formed. A named technical lead has not yet been announced.
Start a conversation
Help us shape a maintenance workflow worth building.
For pilot discussions, operational discovery or company enquiries, contact the team directly.