diff --git a/pipeline.go b/pipeline.go
index 7156065..04ba3f9 100644
--- a/pipeline.go
+++ b/pipeline.go
@@ -1,6 +1,7 @@
 package connector
 
 import (
+	"math"
 	"time"
 
 	"github.com/sendgridlabs/go-kinesis"
@@ -20,6 +21,24 @@ type Pipeline struct {
 	Transformer Transformer
 }
 
+// determine whether the error is recoverable
+func (p Pipeline) isRecoverableError(err error) bool {
+	cErr, ok := err.(*kinesis.Error)
+	if ok && cErr.Code == "ProvisionedThroughputExceeded" {
+		return true
+	}
+	return false
+}
+
+// handle the aws exponential backoff
+// http://docs.aws.amazon.com/general/latest/gr/api-retries.html
+func (p Pipeline) handleAwsWaitTimeExp(attempts int) {
+	// wait up to 5 minutes based on the aws exponential backoff algorithm
+	logger.Printf("waitingnow")
+	waitTime := time.Duration(math.Min(100*math.Pow(2, float64(attempts)), 300000)) * time.Millisecond
+	time.Sleep(waitTime)
+}
+
 // ProcessShard kicks off the process of a Kinesis Shard.
 // It is a long running process that will continue to read from the shard.
 func (p Pipeline) ProcessShard(ksis *kinesis.Kinesis, shardID string) {
@@ -41,6 +60,7 @@ func (p Pipeline) ProcessShard(ksis *kinesis.Kinesis, shardID string) {
 	}
 
 	shardIterator := shardInfo.ShardIterator
+	consecutiveErrorAttempts := 0
 
 	for {
 		args = kinesis.NewArgs()
@@ -48,7 +68,16 @@ func (p Pipeline) ProcessShard(ksis *kinesis.Kinesis, shardID string) {
 		recordSet, err := ksis.GetRecords(args)
 
 		if err != nil {
-			logger.Fatalf("GetRecords ERROR: %v\n", err)
+			if p.isRecoverableError(err) {
+				logger.Printf("GetRecords RECOVERABLE_ERROR: %v\n", err)
+				consecutiveErrorAttempts++
+				p.handleAwsWaitTimeExp(consecutiveErrorAttempts)
+				continue
+			} else {
+				logger.Fatalf("GetRecords ERROR: %v\n", err)
+			}
+		} else {
+			consecutiveErrorAttempts = 0
 		}
 
 		if len(recordSet.Records) > 0 {