Engineer · Builder · Systems thinker
Dependable software,automation & AIsystems for real operations.
I’m Benjamin Mulenga. I design and build intelligent systems that turn complex operations into clarity — spanning software, data, automation, AI and the real operational environments they run in.
Complex operations go in. Useful systems come out.
Production evidence
Built inside live operations, measured there too.
Open any figure for its context, baseline, intervention, my role and what can be verified publicly.
40%lower reporting latency
- Context
- Operational Power BI and SSRS reporting inside a live mining environment.
- Baseline
- The previous reporting path took longer to move production information into decision-ready reports.
- Intervention
- Redesigned the reporting layer and its production data pipelines.
- My role
- Data-system design, implementation, and operational support.
- Verification
- Observed against the previous production reporting path. Internal records and absolute timings are confidential.
Read the system record →15%reduction in equipment idle time
- Context
- Equipment information moving through operational reporting and follow-up workflows.
- Baseline
- Manual and delayed information flows slowed action around idle equipment.
- Intervention
- Introduced Python and SQL automation to improve information speed and consistency.
- My role
- Automation design, data engineering, and delivery.
- Verification
- Reported operational outcome from the improved workflow. Measurement records and attribution detail are not public.
Read the system record →<1hrecovery path, down from four hours
- Context
- Disaster recovery for systems supporting continuous mine operations.
- Baseline
- The documented recovery path was approximately four hours.
- Intervention
- Rebuilt and exercised recovery procedures around operational continuity.
- My role
- Recovery-path design, procedure implementation, and validation.
- Verification
- Documented and exercised recovery path. Infrastructure detail is withheld for security.
Read the system record →5+ yrsinside production operations
- Context
- A 24/7 industrial environment where software and data failures affect physical work.
- Baseline
- Experience spans production IT, data engineering, automation, reporting, and recovery.
- Intervention
- Continuous delivery and support across operational systems rather than a single project.
- My role
- Systems engineer, software developer, data engineer, and automation builder.
- Verification
- Professional operating context; sensitive employer systems and records remain private.
01 — Connect
Fragmented inputsConnected system
Fragmented inputs become one connected system.
Operations already produce the signal — sensors, equipment, databases, spreadsheets, APIs, workflows and the people making calls on shift. The first job is to route it reliably, with its meaning intact.
02 — Process
DataIntelligence
Data becomes intelligence.
A processing layer does the unglamorous work that makes everything downstream trustworthy: ingestion, business rules, automation, AI where it earns its place, validation, and monitoring that makes failure visible before the operation feels it.
03 — Apply
IntelligenceAction
Intelligence becomes action.
Useful systems end in something a person can act on: a dashboard fast enough to change a decision, an alert that reaches the right person, an internal tool, a report, a recommendation, an API another system can trust.
04 — Operate
SoftwareReal-world outcome
Software becomes a real-world outcome.
The loop closes in the physical world — equipment, plant, production, logistics and people. Mining is where I proved it: 24/7 production, physical constraints, recovery, dirty data and human workflows. That systems thinking transfers to any operation that runs on real constraints.
See the systems behind the numbers →05 — Ship
Then the system ships as products.
The same way of working, decomposed into things you can inspect.
