PerfectDraft
65% Analyst Efficiency Gain Through Cloud and Serverless Solutions
-
65%
Increase in analyst efficiency through cloud automation
-
Millions
Returned kegs processed across Europe and the UK
-
10+
Analysts and data scientists moved to real-time cloud tooling
Executive Summary
Primotly engineered a dual-stream cloud transformation and automated logistics platform for PerfectDraft — a global beverage enterprise operating across the UK, Sweden, France, the Netherlands, Spain, Italy, Belgium, the USA, Brazil, and Argentina. The engagement addressed two critical scaling bottlenecks: a legacy, high-maintenance keg return process that required duplicated engineering effort for every new market, and a fragmented data architecture that severely throttled analyst productivity. The solution delivered a standardized, multi-market Keg Return System built on a cost-efficient serverless architecture alongside a modernized Cloud Data Infrastructure spanning Azure and AWS ecosystems. The unified logistics platform has successfully processed millions of returned kegs across Europe and the UK, directly advancing the client's global sustainability goals. Concurrently, the automated data pipeline transitioned over 10 analysts and data scientists to a secure, real-time environment, triggering a 65% increase in analyst efficiency and transforming the unit economics of PerfectDraft's data operations.
About the Client
PerfectDraft is a major international beverage brand operating across four key markets: the USA, UK, France, and Argentina. Specializing in home draft systems, the company manages a complex, high-volume circular economy focused on the distribution, return, and reuse of commercial-grade beverage kegs. With an expanding digital footprint and a rapidly growing customer base, PerfectDraft relies heavily on seamless integration between customer-facing mobile applications, third-party logistics providers, and internal data analysis to maintain operational efficiency and hit strict corporate sustainability targets.
- Custom Software Development
- Cloud & DevOps Engineering
- Data Engineering
Beverage / Logistics / E-commerce
Time & Materials — Dedicated Team
5 people (1 Backend Developer, 1 Data Engineer, 1 DevOps Specialist, 1 Project Manager, 1 Solution Architect)
The Challenge: Legacy Logistics and Data Bottlenecks
As PerfectDraft scaled its international operations, its existing technical infrastructure hit a rigid scaling wall across two distinct operational areas:
- 01/
Inflexible Sustainability Logistics
The core business model relies on a circular economy where customers return empty kegs for reuse. However, the legacy return tracking system was completely unscalable. Every new country rollout or feature adjustment required duplicated development efforts, driving up maintenance costs. PerfectDraft needed a single, unified, cost-effective platform capable of managing international logistics and token-based incentives out of the box.
- 02/
Fragmented Data Architecture
Digital expansion generated massive volumes of data from diverse, siloed sources. The legacy infrastructure lacked real-time processing capabilities, saddling analysts with slow, manual data ingestion tasks. This dependency limited overall team productivity, incurred high cloud operational costs, and left the business blind to real-time operational anomalies.
The Primotly Solution: Serverless Logistics Meets Unified Data
Primotly's architectural strategy attacked both bottlenecks at once by separating the platform into two highly integrated, modern development streams.
The Serverless Keg Platform: Instead of building heavy, monolithic backends for each market, Primotly designed an API-first, serverless keg return platform. By relying on lightweight microservices, the system dynamically scales with multi-country user demand while maintaining near-zero infrastructure overhead during off-peak hours.
The Modern Cloud Data Pipeline: Migrated the client's fragmented data flows into an enterprise-grade cloud ecosystem. By combining the processing power of Azure DataFactory and Databricks with the real-time alerting capabilities of AWS Lambda and Snowflake, the solution eliminated human-bottlenecked manual entry, giving data analysts instant, secure access to clean data.
Development was structured around bi-weekly sprints, utilizing continuous feedback loops and parallel development streams to ensure that new infrastructure layers seamlessly integrated with the logistics engine without interrupting active market operations.
Key Modules Developed
-
Standardized Multi-Market Keg Return Engine
A unified tracking and logistics platform that supports users across multiple European countries and the UK. Built with an API-first design documented via Stoplight, it integrates seamlessly with third-party logistics providers and existing customer apps, replacing high-overhead legacy systems with a single codebase. It has successfully managed the return of millions of kegs and directly supported long-term sustainability goals via integrated token-based incentives.
-
Serverless Infrastructure Integration
Leverages AWS Lambda, SQS, SNS, and CloudWatch to handle real-time data events and messaging queues without managing dedicated virtual servers. Drastically reduces infrastructure run costs by charging only for active execution time, ensuring the return system can handle massive traffic spikes during seasonal promotional waves without risking downtime.
-
Automated Cloud Data Pipeline
A modernized data ingestion and processing layer built using Azure DataFactory, Databricks, Rivery, PySpark, and Python, paired with an enterprise Snowflake data warehouse. Eradicates manual file compilation — over 10 internal analysts and data scientists were transitioned to this automated framework, unlocking a 65% increase in operational efficiency.
-
Real-Time Monitoring & Slack Notification Layer
An automated alerting and dashboard system built via Logic Apps, DataBox, and a custom Slack bot integration that pushes automated PowerBI report snapshots directly to internal communication channels. Converts data monitoring from a passive to a proactive operation, surfacing system anomalies to stakeholders in real time.
-
Enterprise Security & CI/CD Framework
A multi-layered security ecosystem combining Azure Key Vaults, structured logging, Multi-Factor Authentication (MFA) for Snowflake, and automated deployment pipelines managed through Azure DevOps and GitHub Actions. Guarantees enterprise-grade compliance across all active markets, with automated versioning and single-click updates.
Technical Deep Dive
- 01/
Accelerating Multi-Market Rollouts via API-First Serverless Architecture
-
The Problem: The client's legacy return logic was tightly coupled with regional codebases, causing high feature duplication, slow deployments, and unpredictable infrastructure costs as user activity fluctuated across the USA, UK, and Europe.
-
The Solution: Primotly decoupled the core logistics from regional frontends, building a unified serverless API layer. Every service endpoint was rigorously mapped and documented using Stoplight before coding, and the backend logic was written using lightweight Python services triggered on-demand via serverless cloud functions.
-
The Result: infrastructure maintenance costs dropped significantly due to the on-demand scaling model, and the highly structured, API-first blueprint allowed the team to deploy the platform across Europe and the UK rapidly without rewriting core features.
-
- 02/
Orchestrating Multi-Cloud Data Synchronization
-
The Problem: PerfectDraft's operational data sat scattered across separate platforms, restricting analysts to slow, batch-processed insights. The client required a hybrid ecosystem utilizing both Azure data tools and AWS services, but lacked a reliable mechanism to sync real-time data into their Snowflake data warehouse securely.
-
The Solution: Primotly built an automated data bridge. Heavy batch transformation jobs were routed through Azure DataFactory and Databricks using PySpark, while lightweight, time-sensitive events were funneled through AWS Lambda, SQS queues, and Snowpipe directly into Snowflake. The entire environment was locked down using Azure Key Vaults and strict MFA.
-
The Result: human data manipulation was eliminated entirely, and analysts can now query real-time data streams inside Snowflake without waiting on batch execution cycles — driving the 65% efficiency gain.
-
- 03/
Reducing System Downtime with Automated Slack Snapshots
-
The Problem: Analysts and managers were spending excessive time navigating complex monitoring tools and PowerBI dashboards just to verify daily pipeline health and check for logistical errors, leading to delayed responses to data errors.
-
The Solution: Primotly engineered an automated monitoring layer combining Azure Logic Apps and a custom Slack bot. Instead of forcing users to pull data manually, the system captures automated snapshots of targeted PowerBI metrics and operational alerts, pushing them directly into designated Slack channels based on structured logging rules.
-
The Result: manual pipeline monitoring was completely eliminated, and issues are caught and surfaced within minutes of occurrence — allowing the engineering and analytics teams to resolve errors before they impact the broader business.
-
Business Impact & Results
The dual-stream transformation delivered measurable gains across data team velocity, logistics scale, cost efficiency, operational visibility, and corporate sustainability.
-
65% Data Team Velocity Gain
Transitioned 10+ data analysts and scientists away from manual configuration tasks to automated cloud tools, increasing overall data team velocity by 65%.
-
Millions of Kegs Processed
The modernized Keg Return System is fully operational across Europe and the UK, seamlessly tracking millions of returned kegs since launch.
-
Drastic Cost Optimization
The introduction of serverless execution models for the logistics platform successfully lowered base infrastructure running costs.
-
Proactive Operational Visibility
Real-time monitoring, automated cloud alerts, and Slack integrations reduced system downtime and eliminated hours of manual troubleshooting.
-
Corporate Sustainability Alignment
The optimized, token-incentivized return loop directly improved end-customer satisfaction scores while securing PerfectDraft's position as a leader in green, circular-economy logistics.
Tech Stack
- Backend & Logistics
- Python, Serverless Architecture, Stoplight (API Documentation)
- Data Engineering
- Snowflake, Azure DataFactory, Databricks, Rivery, PySpark, PowerBI
- Cloud & Serverless Eventing
- AWS Lambda, Snowpipe, SQS, SNS, Azure Logic Apps
- DevOps & Infrastructure
- Azure DevOps, GitHub Actions, CloudWatch, DataBox, Bash, Azure Key Vaults
- Collaboration & Monitoring
- Slack API (Custom Bot), Jira, Confluence, Google Meet
/ Case studies /
See our case studies
How can we help you?
Let's find the best solution for your business together
Book a meetingHow can we help you?
Let's find the best solution for your business together