endorsedCompleted

AI Image Classification & Microservice for Pynest LTD

Services Covered:

AI and Machine Learning

Industry

Information Technology

Duration

3 months

Budget

10K - 49K

Start Date

1 January 2025

Team Size

1-5 Employees

Engagement Model

Offshore

Project Capability Score (PCS)

80/100

Strong

Project Capability Score (PCS) estimates how capable Yellow Systems 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

Pynest LTD

Cyprus

Project Summary

Yellow built an AI-powered image classification system in Python for Pynest LTD's enterprise client, handling discovery, data preprocessing, model architecture, training, fine-tuning and packaging into a deployable microservice. The model exceeded accuracy targets (over 92%), was delivered on time, and integrated smoothly with the client's backend; communications were proactive and tasks tracked via Slack and Jira.

Key Challenges

Build an AI-powered image classification system in Python for an enterprise client.
In-house team lacked capacity to meet the project's timeline.

Project Deliverables

Discovery phase and data requirements alignment
Data preprocessing and augmentation pipeline
Model architecture design and training
Model fine-tuning and iterative testing
Deployable microservice integrated with client's backend

Project Solution

Yellow ran a discovery phase to align on data requirements and success metrics, designed the model architecture, and set up a data preprocessing and augmentation pipeline. They trained and iteratively tested the model, fine-tuned it for accuracy, and packaged the final model into a deployable microservice that integrated with the client's backend.

Project Outcome

Achieved over 92% accuracy on the client's test dataset.
Delivered the solution on time and integrated smoothly with the client's backend.
Maintained proactive and transparent project management with tracked tasks and early risk flagging.

Platforms

AI/ML Platform

Tech Stack

Python

Client Endorsement

Overall Review Rating

Timeliness

Cost Rating

Willing to Refer

Quality of Deliverables

“The model exceeded accuracy targets, was delivered on time, and integrated smoothly. Communication was proactive and transparent, with tasks tracked and risks flagged early. The team impressed with its technical expertise and flexibility.”
Anastasiya Chyrko
Marketing Manager
Date: 10/21/2025