I advise CTOs, CIOs, platform and data leaders, and investors on data and AI platforms. My services cover performance audits, technical due diligence, capacity planning and platform sizing, resilience and disaster recovery, and technology-specific reviews: Apache Spark, Cassandra, ScyllaDB, MinIO, S3 and Ceph. One method everywhere: measure before concluding, then deliver prioritized risks and decisions you can act on.
My reviews cross every layer, from application code to data models, from processing engines to networking, from object storage to drive specifications and the vendor quote behind the purchase. A diagnosis that stops at a layer boundary relocates the bottleneck instead of removing it. Resilience is validated as engineering, not paperwork: measured restore throughput, replication lag under partition, an RTO and RPO reachability verdict.
Services
- Expert MinIO, S3 and object storage: cluster audits, distributed architecture, multi-site replication, performance and sizing.
- Expert Spark: tuning, benchmark and scale: job measurements, shuffle and partitioning fixes, scale-up plans.
- Expert Apache Cassandra NoSQL: data modeling, p99 latency, repairs, compactions and multi-datacenter design.
- Expert scalability, reliability and disaster recovery: failover testing, measured restore throughput, RTO and RPO reachability verdicts.
- Technical due diligence for data platforms: independent evaluation before a transaction.
- Capacity planning and platform sizing: costed 3 and 5 year growth plans, from workload to cloud instances or a Dell or HPE catalog build.
- Data platform performance audit: latency, throughput, reliability and costs across every layer.
Every engagement ends in a deliverable you can act on: prioritized risks, costed fixes, a sizing plan or negotiation points. No recommendation without a measurement behind it.
FAQ
What services do you offer?
I cover performance audits of data and AI platforms, technical due diligence before a transaction, capacity planning and platform sizing, resilience and disaster recovery with RTO and RPO verdicts, and technology-specific reviews: Spark, Cassandra, ScyllaDB, MinIO, S3 and Ceph. Every engagement delivers prioritized risks and costed fixes.
How does an engagement run?
I start with half a day of access to metrics and teams, measure in production, then deliver a prioritized risk report with costed fixes. Typical duration runs from five days to three weeks depending on scope. A free 15-minute intro call frames which service fits.
Which technologies do you cover?
Data and AI platforms: Apache Spark, Kafka, Kubernetes, PostgreSQL, Cassandra, ScyllaDB, Pulsar, and object storage with MinIO, S3 and Ceph, on public cloud, private cloud or dedicated infrastructure. Twenty years of production work, from banking systems to satellite imagery.