How Ficha Accelerated Iteration and Innovation in WasteTech
From fragmented scripts to structured MLOps: Ficha transformed their workflow to iterate 3x faster on waste classification models.
Get Similar ResultsFicha is an AI-powered waste management company using computer vision in garbage trucks and collection bins to provide cities and housing associations with recycling quality data and insights. Their technology helps municipalities improve recycling rates and reduce contamination.
Ficha was growing rapidly, but their fragmented toolchain couldn't keep up. Despite having access to more data, their models were improving slowly.
Script-based workflows and manual data transfers created significant bottlenecks, preventing the team from moving fast despite rapid company growth.
Organization and tracking of annotations was difficult, making it hard to maintain quality and consistency across the growing dataset.
Model performance improvements lagged despite increasing data availability. Without structured processes, data potential was untapped.
Connecting predictions back to source images and annotations was nearly impossible, making debugging and improvement difficult.
From challenges to solutions
- —Script-based workflows with manual data transfers
- —Fragmented toolchain slowing down the team
- —Difficult annotation organization and tracking
- —Models improving slowly despite data growth
- —No structured process to use available data
- Centralized data visualization and querying
- Structured annotation campaigns
- Integrated experiment tracking with traceability
- End-to-end visibility from predictions to source
- Standardized processes ensuring data reliability
“We suddenly had access to a lot of data, but our models were improving quite slowly. Without a strong process, we couldn't fully benefit from it. Picsellia gave us the structure we needed to iterate faster and keep innovating.”
How Ficha uses Picsellia
Centralize images from garbage trucks and bins into a searchable datalake
Structured campaigns with progress tracking and quality control
Run experiments with full traceability from data to metrics
How Picsellia delivered
Picsellia gave Ficha the structure they needed to iterate faster and keep innovating in waste classification.
Centralized Data Management
Datalake →All waste images centralized with powerful visualization and querying capabilities. No more scattered files or manual transfers.
Structured Annotation Workflows
Annotation Campaigns →Annotation campaigns with clear progress tracking, quality control, and team coordination. 3x improvement in annotation speed.
Full Experiment Traceability
Experiment Tracking →Every experiment tracked with complete lineage. Connect any prediction back to its source images and annotations.
End-to-End Visibility
AI Laboratory →Unified platform replacing fragmented processes. Team can focus on innovation rather than maintenance.
Business impact
With Picsellia, Ficha transformed their waste classification pipeline and accelerated their innovation cycle.
Structured campaigns and better tooling dramatically improved annotation throughput.
Reduced administrative overhead on tracking and management, freeing up time for innovation.
Faster feedback loops between data collection, annotation, training, and deployment.
Reliable, trustworthy AI built on structured workflows and quality-controlled data.
Ready to achieve similar results?
See how Picsellia can help your team build and deploy computer vision models faster.