endorsedOngoingVisual ML Staff Augmentation for Smart Appliance Company
Services Covered:
Industry
Consumer Goods
Duration
107 months
Budget
Confidential
Start Date
1 September 2017
Team Size
1-5 Employees
Engagement Model
Offshore
Project Capability Score (PCS)
74/100
Good
Project Capability Score (PCS) estimates how capable Tooploox 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
Confidential
San Francisco, United States of AmericaProject Summary
Tooploox augmented a smart appliance company's machine-learning team to develop visual-based models that recognize food type and quantity for use in an oven. The vendor worked mostly in Python, building models initially with Caffe and migrating toward TensorFlow Lite. As an augmentation partner since 2017, Tooploox contributed research-focused engineering that improved model breadth, accuracy, and device performance while collaborating closely with the client's in-house team.
Key Challenges
- Difficulty finding machine-learning talent in Silicon Valley
- Need to develop visual ML capable of recognizing type and quantity of food for an oven
- Requirement for research-focused talent and on-device inference optimization
Project Deliverables
- Visual-based machine-learning models to recognize food type and quantity
- Research-based model development implemented in Python
- Caffe-based models and migration/porting toward TensorFlow Lite
Project Solution
Tooploox provided staff augmentation and research-focused machine-learning engineering to build visual-based models that detect food type and estimate quantity for an oven. The team worked primarily in Python, developed models using Caffe, and began migrating the work toward TensorFlow Lite for on-device inference. The engagement involved ongoing collaboration and periodic visits between the client and Tooploox.
Project Outcome
- The models can recognize a broader range of foods
- Model accuracy improved
- Appliance speed and memory usage improved
- Ongoing collaboration with Tooploox acting as a partnering augmentation to the client team
Platforms
Embedded System
AI/ML Platform
Tech Stack
Caffe
TensorFlow
Python
Client Endorsement
Overall Review Rating
4.5
5 out of 5 stars
5 out of 5 starsTimeliness
Cost Rating
Willing to Refer
Quality of Deliverables
“As a result of Tooploox’s work, the client’s machine learning models have improved in breadth, speed, and memory. They act as a true partner, aligning themselves with the client’s goals and working seamlessly with their in-house team.”
Note: This endorsement is based on publicly available client feedback from external review sources.
More Projects by Tooploox
Ongoing
Information Technology
Companion Mobile App Development for Drone Camera Company
Services Covered:
+1 more
Agency name:
Tooploox
Project capability score:
81/100
Ongoing
Information Technology
Staff Augmentation: Three Full-Stack Engineers for Web Platform
Services Covered:
+2 more
Agency name:
Tooploox
Project capability score:
81/100
