Apache Flink SQL, Table API, and UDF development for both OSS Flink and Confluent Cloud
95
95%
Does it follow best practices?
Impact
97%
1.21xAverage score across 5 eval scenarios
Passed
No known issues
Window aggregations and watermarks
Tumbling window TVF
100%
100%
Hopping window TVF
100%
100%
Session window TVF
100%
100%
Cumulate window TVF
100%
100%
Sensor watermark declaration
100%
100%
Activity watermark declaration
100%
100%
Energy watermark declaration
100%
100%
Window GROUP BY includes window_start/end
100%
100%
Kafka connector config
100%
100%
Aggregation functions
100%
100%
DESCRIPTOR references correct column
100%
100%
Partition by sensor/user/meter
100%
100%
UDF development and deployment
Extends ScalarFunction
100%
100%
Public eval method
100%
100%
Null handling in scalar
100%
100%
Extends TableFunction
100%
100%
FunctionHint annotation
100%
100%
Collect method usage
100%
100%
Maven shade plugin
100%
100%
Flink dependency provided scope
100%
100%
Python udf decorator
0%
100%
Artifact upload command
100%
100%
CREATE FUNCTION with artifact
0%
100%
LATERAL TABLE usage
100%
100%
Join patterns and CDC handling
Interval join syntax
100%
100%
Temporal join syntax
100%
100%
Lookup join with PROCTIME
100%
100%
Debezium CDC format
100%
100%
CDC primary key
0%
100%
Upsert-kafka connector
100%
100%
Upsert primary key
100%
100%
Deduplication pattern
100%
100%
Watermarks on streams
100%
100%
Currency rates primary key
100%
100%
State TTL consideration
0%
0%
No unbounded regular join
100%
100%
Process Table Functions
Extends ProcessTableFunction
100%
100%
SET_SEMANTIC_TABLE annotation
0%
100%
StateHint with TTL
20%
100%
ValueState for per-key state
100%
100%
OnTimer annotation
40%
100%
Timer registration
87%
100%
FunctionHint output type
100%
100%
SQL PARTITION BY
100%
100%
SQL uid parameter
0%
100%
SQL on_time DESCRIPTOR
37%
100%
CAST for time parameter
42%
100%
Collect Row output
100%
100%
Confluent Cloud deployment
ConfluentSettings usage
0%
100%
ConfluentTools for results
0%
25%
Statement Set syntax
80%
100%
MATCH_RECOGNIZE structure
100%
100%
System column $rowtime
100%
100%
Compute pool create command
100%
100%
Statement create command
100%
100%
Savepoint workflow
100%
100%
Exception list command
100%
100%
Tumbling window for metrics
100%
100%
Confluent Cloud plugin dependency
25%
100%
No unsupported features
42%
100%