
Paul Chandler
Fractional Apache Cassandra engineer · UK
Cassandra expertise, without the full-time hire.
I’ve spent nearly a decade working with Cassandra at serious scale: running 90+ production clusters for METRO, and providing third-line support for 1,300-node clusters in Ericsson’s billing platform. Now I bring that experience to smaller teams who need an expert, just not a full-time one.
“He is truly an expert in his domain… his ability to navigate complex technical and customer-facing challenges is exceptional.”
Sunil Chadha · Product Owner, Ericsson
1,300
node clusters given third-line support at Ericsson
1,000+
nodes migrated between clouds with zero downtime
50%
storage cut through data model redesign, twice
4.0.3
my patch shipped in Apache Cassandra
Previously trusted by
Ericsson · METRO · Hitachi Consulting · The Workshop · Realistic Games
Sound familiar?
Cassandra is brilliant, until it isn’t.
For teams with one or two clusters, it usually looks like this.
One person understands the cluster and they’re busy, or they’ve left.
Latency creeps up and nobody is quite sure why.
Disk keeps filling up and adding nodes is the only answer anyone has.
You’re stuck on an old version because upgrading feels too risky.
Repairs, compaction, tombstones are on the “we should look at that” list.
No one to call at 3am when it finally goes wrong.
You don’t need a full-time Cassandra engineer. You do need one you can call.
How I help
A fractional Cassandra engineer for your team.
The experience of someone who has run 90+ production clusters and supported some of the largest in the world, for a few hours a month, at a fraction of the cost of a hire.
Start here · Fixed price · Report in 5 working days
Cluster Health Check
I review your cluster end to end: configuration, data model, compaction, repairs, backups, monitoring and capacity. You get a clear written report with prioritised recommendations.
- Written report, prioritised fixes
- Walkthrough call with your team
- Follow-up call after 30 days
- One cluster, up to 12 nodes, one DC
For Hitachi Consulting’s IoT platform, this ended their unplanned outages and let them cut the cluster by 30%.
£3,950
Extra clusters or DCs +£1,250 each · Prices exclude VAT
The full fee is credited against any project or retainer booked within 30 days.
Fixed quote after scoping
Projects
- Version upgrades, including to Cassandra 5.0
- Data model reviews (storage halved for two clients)
- Performance tuning for latency and throughput
- Monitoring with Grafana, Datadog or your stack
- Backup and restore, actually tested
- Migrations: cross-cloud, DSE to Apache
- Automation with Ansible or Puppet
Monthly retainer
Ongoing Support
Your Cassandra engineer on retainer. Proactive monthly checks, repair and patch management, a direct line when something looks wrong, and guaranteed response times when it does.
Essentials
£650 /month
- 4 hours per month
- Monthly health review
- Response within 2 business days
- Unused hours roll over 1 month
Standard Best value
£1,200 /month
- 8 hours per month
- Monthly health review
- Next business day response
- Quarterly upgrade and capacity planning call
- Unused hours roll over 1 month
Response times are the maximum you’ll wait, during UK business hours (Mon–Fri, 9:00–17:30). In practice I’ll often get back to you much sooner. 3-month minimum term, then monthly with 30 days’ notice. Additional hours £150/hour. All prices exclude VAT.
Something’s on fire right now?
Pay-as-you-go incident support for non-clients, subject to availability. I’ve provided third-line support for some of the largest Cassandra clusters in the world.
How it works
Three steps, no hard sell.
01
Free call
I talk through your setup with you and find out what’s worrying you.
02
Health Check
I dig into your cluster and give you a clear picture and a plan.
03
Fix and support
I address the priorities with your team, and you decide whether ongoing support makes sense.
Results
Measured outcomes, not hours billed.
Hitachi Consulting · IoT
Storage halved, outages stopped.
The platform ingested 75 million data points a minute and kept running out of space. Restructuring the data model halved storage, removed the unplanned outages and opened the way to 30% fewer nodes.
METRO · Cloud migration
1,000+ nodes moved, zero downtime.
When Brexit forced ~90 clusters from London to Google Cloud in Belgium, I designed the migration. Every node moved and the data stayed 100% available throughout.
Ericsson · Cassandra 5 upgrade
No critical issues after release.
Shaped the upgrade strategy, found critical risks and bugs early, and drove the fixes through the Apache community before go-live.
Realistic Games · Data model
50% less storage.
A data model review halved their storage requirements, with no extra hardware.
What clients and colleagues say
“He played a key role in shaping our Cassandra 5 upgrade strategy, identifying critical risks and bugs early and driving fixes through the community… thanks to his thorough work, we had no critical issues after release.”
Sunil Chadha · Product Owner, Ericsson
“In a few short months Paul made a huge improvement in the running of our Cassandra cluster. He increased stability, reduced resource consumption and improved monitoring. Would certainly bring Paul back in again.”
Peter Jestico
Technical Design Authority, Hitachi Consulting
“His deep knowledge of Cassandra (both DSE and Apache) made him a key player on our migration between cloud providers, tuning, and cluster split strategy… He also made a huge contribution to our knowledge base.”
Gilberto Müller
Engineering Manager, METRO Deutschland
“A strong, reliable hands-on engineer with deep expertise in Apache Cassandra… proven problem-solving abilities in complex, large-scale systems.”
Ingrid Stanojevic
Technical Product Group Owner, Ericsson
Not just a user. A contributor.
When the problem goes deeper than config, I read the source.
CASSANDRA-17342 Patch accepted into Apache Cassandra 4.0.3, fixing a repair issue
CASSANDRA-20118 Raised and diagnosed bugs in hints and memory handling
CASSANDRA-19785
TCM · 5.1 Tested Transactional Cluster Metadata with the community
ecChronos Features and fixes for the open-source repair scheduler
Speaker DataStax Accelerate, Washington DC: migrating 1,000+ nodes between clouds with zero downtime
Big-estate experience, small-team focus
I’ve seen what goes wrong at 1,300 nodes, so I can spot it early on your six.
Architect as well as operator
A software architecture background means I understand your application, not just the database.
Knowledge transfer built in
I document everything and upskill your team, so you’re never fully dependent on me.
Honest recommendations
I once advised a client not to build a new system, saving an estimated four person-years.
FAQ
Good questions.
We only have one small cluster. Is that too small?
Not at all. Many of the teams I work with run a single production cluster, and that’s often where an outside expert makes the biggest difference.
Do you work remotely?
Yes. I’m UK-based and have worked with clients across the UK and Europe. On-site visits are possible where needed.
Do you support DataStax Enterprise as well as Apache Cassandra?
Yes, both. I’ve also migrated clusters from DSE to open-source Apache Cassandra.
What access do you need?
Often none at all to start with. For a Health Check, you collect your cassandra.yaml and other config files, logs, schema (DDL) and the output of a set of nodetool commands, using a checklist or script I provide, and I analyse them offline. If hands-on access is needed later, we agree it together, whether read-only or supervised sessions. I’m happy to work within your security policies and sign an NDA.
What if I just need a quick question answered?
Book the free call. If it’s a quick fix, I’ll tell you.

Paul Chandler
Apache Cassandra consultant
Stop worrying about your cluster.
Book a free 30-minute call with me and let’s see what’s going on under the hood.