Portcast prediction is powered by proprietary domain-specific Machine Learning algorithms that ingest data from over 70 data sources. By utilizing advanced Artificial Intelligence and orchestrating various data inputs, our system calculates likely arrival and departure delays and explains their reason.
In order to ensure data quality, we have taken the following steps:
- Built validations across all both input data and output data. If there are discrepancies in the data, our system is able to flag the same at scale
- Multiplicity of data: For key data pieces, we get the same information from multiple sources, so that in case of contradictions you can use your judgement and pick the best one
- Built in intelligence: Multiple carriers publish the schedule for every vessel; our system is able to automatically identify the most reliable schedule and uses the same
- Human in the loop: Even after all the above intelligence and automation, some edge cases slip through. Our analysis monitor all outgoing data via various internal dashboards, and step in to make corrections around edge cases
<aside>
☎️ Need further help?
Our friendly Customer Success team is only one message away. Drop the team an email at [email protected] and you’ll get a response within 24-48 hours.
</aside>