Process automation saved substantial cost and increased efficiency of periodical image data processing
Client Background
- An AI solution focussed start-up company based out of UK wanted to develop an employee management application which includes:
- Customer wanted 250 million specious images to be mined. Customer had a list of 2.5 Million species scientific and generic names.
- Customer was particular to collect 2.5 million species all in 10 HD (High Definition) images and sources suggested were Flickr/Instagram/Google.
- Customer was looking for an optimized solution for image data collection and streamline the end-to-end process for the AI solution.
- Also, start-up was looking for a cost-effective solution which can further periodically add to the AI system.
Approach To Solution
- AIMLEAP Team automated the image sourcing process by building. A custom set of web crawlers built including paid Google API to mine image data. The crawler that was able to mine required image data with minimal manual intervention.
- Team performed semi-automated image data validation and ensured that images are of expected quality.
- Also, an automated process established to feed the final set of qualified images to push to customer database on a regular interval.
90%
cost saving
95%
quality assured
250 million images
Quick Turn Around – delivered 250 million images in 60 business days
