Outcomes, not output.
Real engagements where we delivered measurable impact. Some specifics are withheld under NDA — we're happy to talk through detail on a call.
B2B SaaS · Texas
In July of 2024, we reached what felt like the edge of a cliff. All our critical services suddenly came to a grinding halt, and none of us could accurately identify the source of the problem. We knew the problem was related to a particular technology that had been heavily adopted into our code base, so we went out looking for help and expertise on the technology.
We first connected with the team on a Discord forum where we were basically screaming for help. The team replied to my inquiries and first helped with general questions we had, they proved to be very knowledgeable and easy to communicate with, and they quickly gained our trust. So, we decided to bring them on as consultants to help us through the crisis.
Inside of a week, the team was onboarded into our application and was down in the trenches with us troubleshooting the problem. Within 2 weeks we were able to understand the problem, and we found a way to mitigate it so our application could get back to work. But then we went further, and we worked with the team to review our entire application from top to bottom, fleshing out and surfacing other vulnerabilities and potential future crisis.
“The team helped us rapidly get our application back on track, but they also helped us get our hands around the rest of our software to make a road map to put us on the right long-term path. As a manager I felt the team understood our problems and I was never in the dark or out of the loop. As a company we are in a far better place today than before and I would absolutely recommend them to anyone out there needing high level consultations.
Annuities & life insurance · US
Brighthouse first engaged the team in 2020 to design and implement the critical data acquisition and data management capabilities behind a new hedging platform — working directly with their engineers and business partners, from requirements and architecture through to the services that run it.
In 2026 they brought us back to evaluate and modernise their data architecture: analysing existing systems, designing and building a completely new event-driven data management platform, and identifying optimisation opportunities across their wider technology estate.
“What I appreciate most about [the team] is their ability to dive deeply into the problem domain. They work effectively with both our technology teams and our business partners, and they consistently deliver top-level implementations.
We first worked with them on an innovative hedging platform, where they helped us design and implement key data acquisition and management services. The engagement was so successful that we brought them back to help us analyze our data architecture and build a new event-driven data management platform.
They bring strong technical expertise, excellent communication, and great people to work with. I would absolutely work with them again.
Marketplace · an eBay company
Since June 2026 we've been embedded with Depop's Ranking team, supporting the machine learning data pipelines and model-serving infrastructure behind their marketplace.
We led the migration of training-data pipelines to Databricks Asset Bundles — orchestration and application code — and into production, contributed substantially to the move from Jenkins to GitHub Actions, rebuilt the Docker images used to serve models, built a backfill workflow with an at-a-glance progress interface, and helped optimise SageMaker endpoints.
“His work delivered significant performance improvements. Our data-generation jobs now run four to five times faster on average, while model-training jobs run up to twice as fast. Mohammed also developed a backfill workflow with an interface that allows the team to see at a glance what has been processed. Together, these improvements have resolved several persistent pain points and materially improved the team’s day-to-day efficiency.
Beyond his technical contributions, Mohammed integrated quickly and efficiently into the team. He showed strong judgement in identifying where he could add the most value, rather than waiting to be assigned tasks.
We have been extremely pleased with Mohammed’s contribution. Working with Mozek Technologies has been an excellent contractor experience, and we would be very happy to work with them again.
Production ML pipelines
Migrated ranking-model training-data pipelines to Databricks Asset Bundles, moved CI/CD from Jenkins to GitHub Actions and rebuilt model-serving images — making data generation 4–5× faster and training up to 2× faster.
Zero-downtime cloud migration
Migrated a legacy on-prem warehouse to a cloud-native lakehouse with no downtime, cutting run cost while improving query performance.
Production incident turnaround
Embedded within a week to diagnose a platform-wide outage, mitigated it inside two weeks, then reviewed the full application to surface hidden vulnerabilities and set a long-term roadmap.