HOW JARVIS WORKS
You configure Jarvis with read only access to source and target databases. Jarvis can connect to different data sources like Oracle, DB2, SQL server, and can create data models which contains tables, indexes, columns and other database objects to be compared.
Jarvis then shows the source database and the user can select the tables for validation. The data from source is then copied into Jarvis's temporary storage. The same is done for the target database.
Next you export the two databases to Excel. In Excel you can eyeball and write conversion rules to validate if the programmed transformations are correct.
You can then take the excel formula and then Jarvis auto creates the SQL. Jarvis then runs the validation on every record. Giving you 100% coverage and unprecedented confidence.
BENEFITS OF JARVIS
HOW JARVIS SIGNIFICANTLY SAVES COST AND DRAMATICALLY SPEEDS UP DILVERY OF MIGRATION PROJECTS
No Database Corruption
Typically data from source is first exported and imported to target database and then comparison is performed. Which means write operations are happening on production instance. With Jarvis no write happens to any of the databases. The data from the source and target are copied into Jarvis temporary storage which is optimized for validation.
Self-serve
Typically before performing the validation you have to contact the team for the source, and the team for the target and get them to run the conversion rules and then make the source and target availaible for validation. With Jarvis, you directly connect to the source (e.g. DB2) and target database (e.g. SQL server), and Jarvis takes care of the running the conversion rules, extracting the table so you have no dependency on other teams. This dramatically speeds up the validation cycle.
No Performance Impact
Typically, validation are done on the production or dev instances, and that ends up in impacting the performance of the systems. Since Jarvis is a read-only system, the production instances are untouched.
No Cleanup Needed
After the validation, the data imported for validation needs to be deleted. However, this is often forgotten. Unnecessary data in production leads to more maintenance, higher costs, and false alerts. With Jarvis, all these problems are gone because Jarvis keeps the validation data in a separate store from production.
Full Validation Coverage - Each Record (not just eyeball)
Often the business analyst validating the conversion eye balls the results by importing them to excel. They will write an excel formula to get the desired conversion rule and then they just check a few rows. Jarvis has an AI engine which can take the users excel formula, converts that formula to SQL and then runs it on the entire database, and does an exhaustive comparison. So freedom from errors, and validation with confidence.
Faster Validation Cycles
Jarvis uses AI to perform some automated validations. For instance, things like schema, non converted columns, etc are checked giving you peace of mind, and reducing the amount of manual validation