diff --git a/README.md b/README.md
index 1d94ec9..53b8217 100644
--- a/README.md
+++ b/README.md
@@ -65,7 +65,7 @@ situations.
 - Golang 1.14
 
 ## Recent Major Feature Additions/Changes
-- Added `Transform.CurrentRawRecord()` for caller of omniparser to access the raw ingested record.  
+- Added `Transform.RawRecord()` for caller of omniparser to access the raw ingested record.
 - Deprecated `custom_parse` in favor of `custom_func` (`custom_parse` is still usable for
 back-compatibility, it is just removed from all public docs and samples).
 - Added `NonValidatingReader` EDI segment reader.
diff --git a/doc/gettingstarted.md b/doc/gettingstarted.md
index aab50b8..39c1f8b 100644
--- a/doc/gettingstarted.md
+++ b/doc/gettingstarted.md
@@ -715,7 +715,10 @@ for {
         break
     }
     if err != nil { ... }
-    // output contains a []byte of the ingested and transformed record. 
+    // output contains a []byte of the ingested and transformed record.
+
+    // Also transform.RawRecord() gives you access to the raw record.
+    fmt.Println(transform.RawRecord().Checksum())
 }
 ```
 
diff --git a/doc/programmability.md b/doc/programmability.md
index f170b24..29d4d4f 100644
--- a/doc/programmability.md
+++ b/doc/programmability.md
@@ -35,11 +35,9 @@ for {
     }
     if err != nil { ... }
     // output contains a []byte of the ingested and transformed record.
-    
-    raw, err := transform.CurrentRawRecord()
-    if err != nil { ... }
-    rawRecord := raw.(*omniv21.RawRecord) // assuming the schema is of `omni.2.1` version.
-    fmt.Println(rawRecord.UUIDv3()) // rawRecord.UUIDv3() returns a stable hash of the current raw record. 
+
+    // Also transform.RawRecord() gives you access to the raw record.
+    fmt.Println(transform.RawRecord().Checksum())
 }
 ```
 Note this out-of-box omniparser setup contains only the `omni.2.1` schema handler, meaning only schemas
diff --git a/extensions/omniv21/ingester.go b/extensions/omniv21/ingester.go
index 3b13c22..0e0f9c6 100644
--- a/extensions/omniv21/ingester.go
+++ b/extensions/omniv21/ingester.go
@@ -9,19 +9,21 @@ import (
 	"github.com/jf-tech/omniparser/extensions/omniv21/fileformat"
 	"github.com/jf-tech/omniparser/extensions/omniv21/transform"
 	"github.com/jf-tech/omniparser/idr"
+	"github.com/jf-tech/omniparser/schemahandler"
 	"github.com/jf-tech/omniparser/transformctx"
 )
 
-// RawRecord contains the raw data ingested in from the input stream in the form of an IDR tree.
-// Note callers outside this package should absolutely make **NO** modifications to the content of
-// RawRecord. Treat it like read-only.
-type RawRecord struct {
-	Node *idr.Node
+type rawRecord struct {
+	node *idr.Node
 }
 
-// UUIDv3 returns a stable MD5(v3) hash of the RawRecord.
-func (rr *RawRecord) UUIDv3() string {
-	hash, _ := customfuncs.UUIDv3(nil, idr.JSONify2(rr.Node))
+func (rr *rawRecord) Raw() interface{} {
+	return rr.node
+}
+
+// Checksum returns a stable MD5(v3) hash of the rawRecord.
+func (rr *rawRecord) Checksum() string {
+	hash, _ := customfuncs.UUIDv3(nil, idr.JSONify2(rr.node))
 	return hash
 }
 
@@ -31,19 +33,19 @@ type ingester struct {
 	customParseFuncs transform.CustomParseFuncs // Deprecated.
 	ctx              *transformctx.Ctx
 	reader           fileformat.FormatReader
-	rawRecord        RawRecord
+	rawRecord        rawRecord
 }
 
 // Read ingests a raw record from the input stream, transforms it according the given schema and return
 // the raw record, transformed JSON bytes.
-func (g *ingester) Read() (interface{}, []byte, error) {
-	if g.rawRecord.Node != nil {
-		g.reader.Release(g.rawRecord.Node)
-		g.rawRecord.Node = nil
+func (g *ingester) Read() (schemahandler.RawRecord, []byte, error) {
+	if g.rawRecord.node != nil {
+		g.reader.Release(g.rawRecord.node)
+		g.rawRecord.node = nil
 	}
 	n, err := g.reader.Read()
 	if n != nil {
-		g.rawRecord.Node = n
+		g.rawRecord.node = n
 	}
 	if err != nil {
 		// Read() supposed to have already done CtxAwareErr error wrapping. So directly return.
diff --git a/extensions/omniv21/ingester_test.go b/extensions/omniv21/ingester_test.go
index ccf365d..273e23b 100644
--- a/extensions/omniv21/ingester_test.go
+++ b/extensions/omniv21/ingester_test.go
@@ -91,7 +91,8 @@ func TestIngester_Read_Success(t *testing.T) {
 	}
 	raw, b, err := g.Read()
 	assert.NoError(t, err)
-	assert.Equal(t, "41665284-dab9-300d-b647-7ace9cb514b4", raw.(*RawRecord).UUIDv3())
+	assert.Equal(t, "41665284-dab9-300d-b647-7ace9cb514b4", raw.Checksum())
+	assert.Equal(t, "{}", idr.JSONify2(raw.Raw().(*idr.Node)))
 	assert.Equal(t, "123", string(b))
 	assert.Equal(t, 0, g.reader.(*testReader).releaseCalled)
 	raw, b, err = g.Read()
diff --git a/extensions/omniv21/samples/testCommon.go b/extensions/omniv21/samples/testCommon.go
index f8abfe2..139de73 100644
--- a/extensions/omniv21/samples/testCommon.go
+++ b/extensions/omniv21/samples/testCommon.go
@@ -11,7 +11,6 @@ import (
 	"github.com/stretchr/testify/assert"
 
 	"github.com/jf-tech/omniparser"
-	"github.com/jf-tech/omniparser/extensions/omniv21"
 	"github.com/jf-tech/omniparser/idr"
 	"github.com/jf-tech/omniparser/transformctx"
 )
@@ -49,12 +48,11 @@ func SampleTestCommon(t *testing.T, schemaFile, inputFile string) string {
 		err = json.Unmarshal(recordBytes, &transformed)
 		assert.NoError(t, err)
 
-		raw, err := transform.CurrentRawRecord()
+		raw, err := transform.RawRecord()
 		assert.NoError(t, err)
-		rawRecord := raw.(*omniv21.RawRecord)
 		records = append(records, record{
-			RawRecord:         idr.JSONify2(rawRecord.Node),
-			RawRecordHash:     rawRecord.UUIDv3(),
+			RawRecord:         idr.JSONify2(raw.Raw().(*idr.Node)),
+			RawRecordHash:     raw.Checksum(),
 			TransformedRecord: transformed,
 		})
 	}
diff --git a/schemahandler/schemaHandler.go b/schemahandler/schemaHandler.go
index 897b800..bdb43de 100644
--- a/schemahandler/schemaHandler.go
+++ b/schemahandler/schemaHandler.go
@@ -37,6 +37,14 @@ type SchemaHandler interface {
 	NewIngester(ctx *transformctx.Ctx, input io.Reader) (Ingester, error)
 }
 
+// RawRecord represents a raw record ingested from the input.
+type RawRecord interface {
+	// Raw returns the actual raw record that is version specific to each of the schema handler.
+	Raw() interface{}
+	// Checksum returns a UUIDv3 (MD5) stable hash of the raw record.
+	Checksum() string
+}
+
 // Ingester is an interface of ingestion and transformation for a given input stream.
 type Ingester interface {
 	// Read is called repeatedly during the processing of an input stream. Each call it should return
@@ -46,7 +54,7 @@ type Ingester interface {
 	// one record at a time, OR, processes and returns one record for each call. However, the overall
 	// design principle of omniparser is to have streaming processing capability so memory won't be a
 	// constraint when dealing with large input file. All built-in ingesters are implemented this way.
-	Read() (interface{}, []byte, error)
+	Read() (RawRecord, []byte, error)
 
 	// IsContinuableError is called to determine if the error returned by Read is fatal or not. After Read
 	// is called, the result record or error will be returned to caller. After caller consumes record or
diff --git a/transform.go b/transform.go
index 96c5443..f5c32b4 100644
--- a/transform.go
+++ b/transform.go
@@ -21,16 +21,14 @@ type Transform interface {
 	// return the same error.
 	// Note if returned error isn't nil, then returned []byte will be nil.
 	Read() ([]byte, error)
-	// CurrentRawRecord returns the current raw record ingested from the input stream. If
-	// the last Read call failed, or Read hasn't been called yet, it will return an error.
-	// Each schema handler and extension has its own definition of what a raw record is
-	// so please check their corresponding doc.
-	CurrentRawRecord() (interface{}, error)
+	// RawRecord returns the current raw record ingested from the input stream. If the last
+	// Read call failed, or Read hasn't been called yet, it will return an error.
+	RawRecord() (schemahandler.RawRecord, error)
 }
 
 type transform struct {
 	ingester      schemahandler.Ingester
-	lastRawRecord interface{}
+	lastRawRecord schemahandler.RawRecord
 	lastErr       error
 }
 
@@ -70,9 +68,9 @@ func (o *transform) Read() ([]byte, error) {
 	return transformed, err
 }
 
-// CurrentRawRecord returns the current raw record ingested from the input stream. If
-// the last Read call failed, or Read hasn't been called yet, it will return an error.
-func (o *transform) CurrentRawRecord() (interface{}, error) {
+// RawRecord returns the current raw record ingested from the input stream. If the last
+// Read call failed, or Read hasn't been called yet, it will return an error.
+func (o *transform) RawRecord() (schemahandler.RawRecord, error) {
 	if o.lastErr != nil {
 		return nil, o.lastErr
 	}
diff --git a/transform_test.go b/transform_test.go
index e47ca62..482beab 100644
--- a/transform_test.go
+++ b/transform_test.go
@@ -9,26 +9,41 @@ import (
 	"github.com/stretchr/testify/assert"
 
 	"github.com/jf-tech/omniparser/errs"
+	"github.com/jf-tech/omniparser/schemahandler"
 )
 
 type testReadCall struct {
-	record []byte
+	result []byte
 	err    error
 }
 
+func (trc testReadCall) Checksum() string {
+	if trc.err != nil {
+		panic("Checksum() called when err != nil")
+	}
+	return fmt.Sprintf("checksum of raw record of '%s'", string(trc.result))
+}
+
+func (trc testReadCall) Raw() interface{} {
+	if trc.err != nil {
+		panic("Raw() called when err != nil")
+	}
+	return fmt.Sprintf("raw record of '%s'", string(trc.result))
+}
+
 type testIngester struct {
 	readCalled      int
 	readCalls       []testReadCall
 	continuableErrs map[error]bool
 }
 
-func (g *testIngester) Read() (interface{}, []byte, error) {
+func (g *testIngester) Read() (schemahandler.RawRecord, []byte, error) {
 	if g.readCalled >= len(g.readCalls) {
 		panic(fmt.Sprintf("Read() called %d time(s), but not enough mock entries setup", g.readCalled))
 	}
 	r := g.readCalls[g.readCalled]
 	g.readCalled++
-	return fmt.Sprintf("raw record %d", g.readCalled-1), r.record, r.err
+	return r, r.result, r.err
 }
 
 func (g *testIngester) IsContinuableError(err error) bool {
@@ -45,9 +60,9 @@ func TestTransform_Read_EndWithEOF(t *testing.T) {
 	tfm := &transform{
 		ingester: &testIngester{
 			readCalls: []testReadCall{
-				{record: []byte("1st good read")},
+				{result: []byte("1st good read")},
 				{err: continuableErr1},
-				{record: []byte("2nd good read")},
+				{result: []byte("2nd good read")},
 				{err: io.EOF},
 			},
 			continuableErrs: map[error]bool{continuableErr1: true},
@@ -56,16 +71,17 @@ func TestTransform_Read_EndWithEOF(t *testing.T) {
 	record, err := tfm.Read()
 	assert.NoError(t, err)
 	assert.Equal(t, "1st good read", string(record))
-	raw, err := tfm.CurrentRawRecord()
+	raw, err := tfm.RawRecord()
 	assert.NoError(t, err)
-	assert.Equal(t, "raw record 0", raw.(string))
+	assert.Equal(t, "raw record of '1st good read'", raw.Raw())
+	assert.Equal(t, "checksum of raw record of '1st good read'", raw.Checksum())
 
 	record, err = tfm.Read()
 	assert.Error(t, err)
 	assert.True(t, errs.IsErrTransformFailed(err))
 	assert.Equal(t, continuableErr1.Error(), err.Error())
 	assert.Nil(t, record)
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.Error(t, err)
 	assert.True(t, errs.IsErrTransformFailed(err))
 	assert.Nil(t, raw)
@@ -73,15 +89,16 @@ func TestTransform_Read_EndWithEOF(t *testing.T) {
 	record, err = tfm.Read()
 	assert.NoError(t, err)
 	assert.Equal(t, "2nd good read", string(record))
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.NoError(t, err)
-	assert.Equal(t, "raw record 2", raw.(string))
+	assert.Equal(t, "raw record of '2nd good read'", raw.Raw())
+	assert.Equal(t, "checksum of raw record of '2nd good read'", raw.Checksum())
 
 	record, err = tfm.Read()
 	assert.Error(t, err)
 	assert.Equal(t, io.EOF, err)
 	assert.Nil(t, record)
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.Error(t, err)
 	assert.Equal(t, io.EOF, err)
 	assert.Nil(t, raw)
@@ -91,7 +108,7 @@ func TestTransform_Read_EndWithEOF(t *testing.T) {
 	assert.Error(t, err)
 	assert.Equal(t, io.EOF, err)
 	assert.Nil(t, record)
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.Error(t, err)
 	assert.Equal(t, io.EOF, err)
 	assert.Nil(t, raw)
@@ -101,7 +118,7 @@ func TestTransform_Read_EndWithNonContinuableError(t *testing.T) {
 	tfm := &transform{
 		ingester: &testIngester{
 			readCalls: []testReadCall{
-				{record: []byte("1st good read")},
+				{result: []byte("1st good read")},
 				{err: errors.New("fatal error")},
 			},
 		},
@@ -109,16 +126,17 @@ func TestTransform_Read_EndWithNonContinuableError(t *testing.T) {
 	record, err := tfm.Read()
 	assert.NoError(t, err)
 	assert.Equal(t, "1st good read", string(record))
-	raw, err := tfm.CurrentRawRecord()
+	raw, err := tfm.RawRecord()
 	assert.NoError(t, err)
-	assert.Equal(t, "raw record 0", raw.(string))
+	assert.Equal(t, "raw record of '1st good read'", raw.Raw())
+	assert.Equal(t, "checksum of raw record of '1st good read'", raw.Checksum())
 
 	record, err = tfm.Read()
 	assert.Error(t, err)
 	assert.False(t, errs.IsErrTransformFailed(err))
 	assert.Equal(t, "fatal error", err.Error())
 	assert.Nil(t, record)
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.Error(t, err)
 	assert.False(t, errs.IsErrTransformFailed(err))
 	assert.Equal(t, "fatal error", err.Error())
@@ -129,15 +147,15 @@ func TestTransform_Read_EndWithNonContinuableError(t *testing.T) {
 	assert.Error(t, err)
 	assert.Equal(t, "fatal error", err.Error())
 	assert.Nil(t, record)
-	raw, err = tfm.CurrentRawRecord()
+	raw, err = tfm.RawRecord()
 	assert.Error(t, err)
 	assert.Equal(t, "fatal error", err.Error())
 	assert.Nil(t, raw)
 }
 
-func TestTransform_CurrentRawRecord_CalledBeforeRead(t *testing.T) {
+func TestTransform_RawRecord_CalledBeforeRead(t *testing.T) {
 	tfm := &transform{ingester: &testIngester{readCalls: []testReadCall{}}}
-	raw, err := tfm.CurrentRawRecord()
+	raw, err := tfm.RawRecord()
 	assert.Error(t, err)
 	assert.Equal(t, "must call Read first", err.Error())
 	assert.Nil(t, raw)