Our work - Proven solutions for real-world problems.

We believe in efficiency and maximizing our resources to provide the best value to our clients. The primary way we do that is by re-using the same five projects we’ve been developing for the past decade.

Case studies

Cadogan Tate

Search and retrieval, data engineering, computer vision, iOS

Making a warehouse of unlabelled art searchable

Cadogan Tate stores and moves irreplaceable art across thirteen locations in the UK, Europe and the US. Every record sat in a dated ERP with no web interface, so clients asked account managers where their items were and waited days for a reply.

We built a sub-minute read path out of the ERP, a search-driven client portal, and an iOS app that builds packing manifests from photographs — without replacing the system that runs the business.

They consistently delivered high-quality work on aggressive timelines without compromising design or performance. It is rare to find a team that blends speed, reliability, and polish so seamlessly.

Mark Isitt, CTO of Cadogan Tate

Atlassian

Search infrastructure, security engineering

Customer-held encryption keys on a shared search cluster

Regulated-sector customers were asking for full cryptographic ownership of their data. Atlassian search runs on multi-tenant Elasticsearch, encrypted at rest with platform-managed keys, so there was no obvious place for a customer key to live.

We built a working proof of concept: envelope encryption on the write path, root keys held in the customer account, and a toggle at cluster level so nothing above the cluster had to change.

Despite the complexity and sensitivity of the task, they ramped up quickly, worked closely with our security and infra stakeholders, and delivered a fully functional proof of concept that met all our requirements.

Namrata Lele, Head of Engineering, Atlassian

BILL

Data engineering, cloud migration

Data lake migration takes dashboards from hours behind to minutes

BILL ran analytics for its payables and receivables platform on a legacy on-prem cluster fed by Oracle and SQL Server ETLs. Data growth and query concurrency had outpaced it, dashboards ran minutes slow at peak, and new capacity meant month-long hardware procurement.

We moved 200 TB to an S3-backed lake with continuous change-data-capture replication, verified parity by checksum rather than by assertion, and did it without pausing financial close reporting.

Levi's

Search and recommendations, cloud migration, data engineering

A recommender that went from three seconds to under a hundred milliseconds

Levi's ran its site on SAP Hybris with nightly cron ETLs that had no access to session data. Finance and merchandising waited a day or more for performance reports, holiday peaks saturated on-prem capacity, and the recommender took three seconds to answer while blind to what was actually in stock.

We lifted the web tier to the cloud, centralised click-level and order data in one lake, and rebuilt recommendations as a vector search problem — which is what it always was.

Sidebetz

Game development

Fantasy league platform grows revenue by 85%

Enabled a gaming industry leader to unlock a new revenue stream from live games by leveraging Vision AI for real-time game insights, driving user speculation.

Our solution collects live streams of games from multiple locations and extracts information using Object Detection and OCR.

The two-person team operated with clarity, autonomy, and rigor, matching the standards we expect internally at Atlassian. It was plug-and-play execution with zero overhead on our end—exactly what we needed.

Namrata Lele, Head of Engineering, Atlassian

AtlassianAtlassian colored

You’re in good company

  • atlassianatlassian
  • alibabagroupalibabagroup
  • levislevis
  • billbill
  • cadogantatecadogantate
  • 3m3m
  • xdc-networkxdc-network
  • sidebetzsidebetz

Tell us about your project

Our offices

  • USA
    Palo Alto, CA
    Oakland, CA
  • INDIA
    Vadodara, GJ