diff --git a/internal/observability/tracer.go b/internal/observability/tracer.go
index 25bd6a2..9b0fb53 100644
--- a/internal/observability/tracer.go
+++ b/internal/observability/tracer.go
@@ -2,6 +2,10 @@ package observability
 
 import (
 	"context"
+	"time"
+
+	"google.golang.org/grpc"
+	"google.golang.org/grpc/backoff"
 
 	"go.opentelemetry.io/otel"
 	"go.opentelemetry.io/otel/attribute"
@@ -9,10 +13,9 @@ import (
 	"go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc"
 	"go.opentelemetry.io/otel/sdk/resource"
 	"go.opentelemetry.io/otel/sdk/trace"
-	"go.opentelemetry.io/otel/trace/noop"
-
 	semconv "go.opentelemetry.io/otel/semconv/v1.26.0"
 	oteltrace "go.opentelemetry.io/otel/trace"
+	"go.opentelemetry.io/otel/trace/noop"
 )
 
 type TracerOptions struct {
@@ -33,6 +36,27 @@ func SetupTracer(opts TracerOptions) (func() error, error) {
 	otlpClient := otlptracegrpc.NewClient(
 		otlptracegrpc.WithInsecure(),
 		otlptracegrpc.WithEndpoint(opts.OTLPEndpoint),
+		otlptracegrpc.WithTimeout(90*time.Second),
+		// Configure gRPC connection retry for network-level failures (e.g., "connection refused")
+		otlptracegrpc.WithDialOption(
+			grpc.WithConnectParams(grpc.ConnectParams{
+				Backoff: backoff.Config{
+					// this delay/multiplier config gives retries at roughly [3, 9, 21, 45, 93] seconds
+					BaseDelay:  3 * time.Second,
+					Multiplier: 2.0,
+					Jitter:     0.2,
+					MaxDelay:   120 * time.Second,
+				},
+				MinConnectTimeout: 5 * time.Second,
+			}),
+		),
+		// Configure application-level retry for retryable errors (e.g., rate limits, temporary server errors)
+		otlptracegrpc.WithRetry(otlptracegrpc.RetryConfig{
+			Enabled:         true,
+			InitialInterval: 5 * time.Second,
+			MaxInterval:     15 * time.Second,
+			MaxElapsedTime:  90 * time.Second,
+		}),
 	)
 
 	exporter, err := otlptrace.New(context.Background(), otlpClient)