VoidmodeStudios
Portfolio · Not taking on workSee the work

Enterprise work, anonymized

The bigger systems, and what they changed.

The problem, the result and the engineering behind it. Client names are anonymized where confidentiality requires it.

Events a year
Uptime delivered
Annual cost removed

Enterprise

Three that moved a real number.

  1. Telecommunications events platform

    Major UK telecoms provider, anonymized · 7 billion events a year

    The problem

    A legacy system buckling under exponential data growth, with enterprise ERP integration requirements layered on top and no tolerance for downtime.

    The approach

    A complete microservices redesign with fault-tolerant event processing, migrated piece by piece so live traffic never noticed.

    1000x
    Event processing speed
    99.99%
    Uptime achieved
    £2.5M
    Annual cost reduction
    1,000s/s
    Peak events per second
    Lead systems architect and implementation lead18 monthsMicroservices · Event streaming · ERP integration · Fault tolerance · High availability

  2. Real-time analytics transformation

    Enterprise software company, anonymized · Multi-tenant SaaS platform

    The problem

    Analytics queries taking 45 seconds or more, with customer churn climbing because of it.

    The approach

    Event-driven architecture with optimized data pipelines and caching strategies aimed at the queries people actually ran.

    <1s
    Query response
    -23%
    Customer churn
    £800K
    ARR retained
    -60%
    Infrastructure cost
    Performance engineering lead8 monthsEvent streaming · Pipeline optimization · Caching · Performance tuning

  3. Critical system recovery

    Financial services provider, anonymized · Multi-million transaction platform

    The problem

    Complete system failure during peak trading hours, losing roughly £50,000 a minute while it was down.

    The approach

    A 72-hour emergency architecture rebuild with a zero-downtime deployment strategy, then a proper rebuild behind it.

    72 hours
    Full restoration
    0
    Critical failures since
    £2.1M
    Downtime prevented
    Built in
    Disaster recovery
    Crisis response technical lead3 days to recovery, 3 months to full rebuildDisaster recovery · Zero-downtime deployment · System resilience · Emergency response

Anonymized where client confidentiality requires it.

Named work

The ones we can put a name to.

  1. mitchs.aiA production platform generating personalized ads at 92% accuracy, with users reporting 70% time savings against writing by hand. Revenue has climbed every month since launch in July 2025.Recruitment · AI
  2. Trap Music Museum47% revenue increase in the first quarter, manual work down 60%, and mobile conversion doubled. Zero inventory sync errors since launch.Entertainment · Commerce
  3. WideZike15K+ beta users in the first month, 67% engagement, and 18-minute average sessions. The recommendation engine reached 92% accuracy on content suggestions.Social video platform
  4. The Sound Learning Centre210% increase in mobile traffic, a 95/100 PageSpeed score, and content updates that take minutes rather than hours.Education platform
  5. Frestons180% increase in organic traffic, top three Google rankings on key terms, and the ongoing hosting and maintenance cost removed entirely.Property · SEO

Strictly a portfolio, for now.

I'm not taking on client work right now, so this site is just here to show what's been built. The full portfolio has games you can play, products you can use and the stories behind the bigger systems.

See the work Our own products