endorsedCompleted

Plugin Ecosystem Development for Time-Series Platform

Services Covered:

Custom Software Development • Test Automation • DevOps • AI and Machine Learning

Industry

Software Development

Duration

19 months

Budget

50K - 199K

Start Date

1 February 2024

Team Size

1-5 Employees

Engagement Model

Offshore

Project Capability Score (PCS)

84/100

Strong

Project Capability Score (PCS) estimates how capable Pynest is of successfully delivering a project like this, based on its past experience, track record, and ability to handle similar work. Learn more about PCS

Client

Yellow Systems

United States of America

Project Summary

Pynest provided custom software development services to build a plugin ecosystem for a high-performance time-series data platform, delivering over 15 plugins and extended SDK modules. The engagement included testing and automation, CI/CD workflow management, and integrations with LLM tooling. The team communicated via GitHub and Slack and consistently delivered stable, production-ready components on time.

Key Challenges

Build a plugin ecosystem for a high-performance time-series data platform
Develop 15+ plugins for data transformation, replication, forecasting, anomaly detection, and real-time notifications
Contribute to SDK development and integration testing for LLM-based tools
Automate internal AI workflows

Project Deliverables

Plugin development (15+ plugins)
Extended SDK modules
Automated test suites
CI/CD (GitHub Actions) workflows
Internal demo apps using Writer SDK and Agent Builder
Real-time notification plugins (HTTP, Slack, Discord, SMS, WhatsApp)
Anomaly detection plugins (MAD check, deadman check, state change detection, threshold check, forecasting with Prophet)
Data transformation and replication plugins

Project Solution

Pynest provided software development services for a high-performance time-series data platform, delivering plugin development for data analytics and configuration, extended SDK module development, testing and automation, and CI/CD workflow management. Engineers implemented downsampling with configurable aggregations, data transformation and normalization (Pint), anomaly detection plugins, real-time notification plugins with multiple providers, data replication plugins, and extended SDK/AI tooling using LangChain and AWS Bedrock. They also built internal demo apps, automated test suites (Pytest, UnitTest, Playwright), and GitHub Actions workflows for CI/CD.

Project Outcome

Delivery of 15+ stable, modular, production-ready plugins
Improved observability and forecasting within the time-series platform
Streamlined internal workflows via SDK integration
Faster iteration speeds for LLM-based features
Consistent on-time delivery and clear communications via GitHub and Slack

Platforms

Cloud

Tech Stack

Python
Amazon Web Services (AWS)
Pytest

Client Endorsement

Overall Review Rating

Timeliness

Cost Rating

Willing to Refer

Quality of Deliverables

“Pynest successfully delivered stable, modular, production-ready plug-ins that improved observability and forecasting within the client's platform. The team communicated regularly through GitHub and Slack and consistently submitted deliverables on time. Their expertise and knowledge also stood out.”
Mitya Smusin
CEO
Date: 10/20/2025