diff --git a/rust/src/index/vector/ivf.rs b/rust/src/index/vector/ivf.rs
index 8343ca2d17..5279845d2b 100644
--- a/rust/src/index/vector/ivf.rs
+++ b/rust/src/index/vector/ivf.rs
@@ -790,9 +790,11 @@ impl IndexBuilder for IvfPqIndexBuilder {
             ProductQuantizer::new(self.num_sub_vectors as usize, self.nbits, self.dimension);
         let batch = concat_batches(&partitioned_batches[0].schema(), &partitioned_batches)?;
         let residual_vector = batch.column_by_name(RESIDUAL_COLUMN).unwrap();
+        let data: &Float32Array = as_primitive_array(residual_vector);
+        let resid_mat = MatrixView::new(Arc::new(data.clone()), self.dimension);
 
         let pq_code = pq
-            .fit_transform(as_fixed_size_list_array(residual_vector), self.metric_type)
+            .fit_transform(&resid_mat, self.metric_type)
             .await?;
 
         const PQ_CODE_COLUMN: &str = "__pq_code";
diff --git a/rust/src/index/vector/opq.rs b/rust/src/index/vector/opq.rs
index 0bea944a2d..b1ff8ed0f3 100644
--- a/rust/src/index/vector/opq.rs
+++ b/rust/src/index/vector/opq.rs
@@ -76,11 +76,8 @@ impl OptimizedProductQuantizer {
         train: &MatrixView,
         metric_type: MetricType,
     ) -> Result<(MatrixView, ProductQuantizer)> {
-        let dim = train.num_columns();
-        // TODO: make PQ::fit_transform work with MatrixView.
-        let fixed_list = FixedSizeListArray::try_new(train.data().as_ref(), dim as i32)?;
         let mut pq = ProductQuantizer::new(self.num_sub_vectors, self.num_bits, dim);
-        let pq_code = pq.fit_transform(&fixed_list, metric_type).await?;
+        let pq_code = pq.fit_transform(&train, metric_type).await?;
 
         // Reconstruct Y
         let mut builder = Float32Builder::with_capacity(train.num_columns() * train.num_rows());
diff --git a/rust/src/index/vector/pq.rs b/rust/src/index/vector/pq.rs
index 6f7ac5b2d5..2438ecc4c8 100644
--- a/rust/src/index/vector/pq.rs
+++ b/rust/src/index/vector/pq.rs
@@ -26,6 +26,7 @@ use futures::stream::{self, StreamExt, TryStreamExt};
 use rand::SeedableRng;
 
 use crate::arrow::*;
+use crate::arrow::linalg::MatrixView;
 use crate::index::pb;
 use crate::index::vector::kmeans::train_kmeans;
 use crate::io::object_reader::{read_fixed_stride_array, ObjectReader};
@@ -400,12 +401,13 @@ impl ProductQuantizer {
     /// Train a [ProductQuantizer] using an array of vectors.
     pub async fn fit_transform(
         &mut self,
-        data: &FixedSizeListArray,
+        mat: &MatrixView,
         metric_type: MetricType,
     ) -> Result<FixedSizeListArray> {
-        self.train(data, metric_type, 50).await?;
+        let data = FixedSizeListArray::try_new(mat.data().as_ref(), mat.num_columns() as i32)?;
+        self.train(&data, metric_type, 50).await?;
 
-        let sub_vectors = divide_to_subvectors(data, self.num_sub_vectors as i32);
+        let sub_vectors = divide_to_subvectors(&data, self.num_sub_vectors as i32);
         self.transform(&sub_vectors, metric_type).await
     }
 }
@@ -430,7 +432,6 @@ impl From<&ProductQuantizer> for pb::Pq {
             num_sub_vectors: pq.num_sub_vectors as u32,
             dimension: pq.dimension as u32,
             codebook: pq.codebook.as_ref().unwrap().values().to_vec(),
-            opq: None,
         }
     }
 }