Local sample libraries, finally searchable

Find the right sound before the idea disappears.

Scan local sample folders, classify sounds with AI, extract BPM, key, and duration, then search thousands of loops, one-shots, drums, vocals, FX, and instruments from one producer-focused library.

LOCAL INDEXINGOriginal files stay putAUDIO-AWARE AISearch by how it soundsPRODUCER METADATABPM, key, type, and tags
SAMPLEFIND
K
Search
YOUR SOUND LIBRARYYour sample library, organized.

Search, filter, favorite, and preview every local sound.

42 results matched “grand piano melody”
Concert_Grand_Arpeggio_Gm_90.wavConcert grand · flowing arpeggio
PianoLoop
0:16Gm90
Felt_Grand_Piano_Motif_Gm_94.wavSoft felt piano · melodic motif
PianoLoop
0:11Gm94
Warm_Grand_Chords_Am_92.wavWarm grand piano · chord phrase
PianoLoop
0:12Am92
Classical_Grand_Phrase_Fm_96.wavClassical grand · lyrical phrase
PianoLoop
0:15Fm96
Grand_Piano_Melody_Cm_92.wav
0:070:13
ONE LIBRARY · ANY SOUND · ZERO FOLDER DIVING
SEARCH · PREVIEW · CREATE
THE PROBLEM

Your best sounds shouldn’t be buried in folders.

Sample libraries grow fast. Filenames become inconsistent, categories overlap, and auditioning folders one by one interrupts the moment when an idea is taking shape.

SampleFind turns scattered audio into one intelligent, searchable production library—without moving or uploading the original files.

01Understand

Search how producers think

Describe a sound, name an instrument, enter a key or BPM, or combine them. SampleFind turns producer language into focused audio results.

“dark darbuka loop in A minor”
02Organize

Turn folders into a library

Index the folders you choose while your original audio stays exactly where it is. Sources remain manageable, rescannable, and removable.

Local files. Structured metadata.
03Create

Move from search to session

Preview sounds instantly, favorite the best matches, then drag audio directly into supported DAWs without breaking your creative flow.

Find → preview → drag into DAW
04Control

Filter like a producer

Layer sample type, category, key, BPM, and favorites to narrow a large library without losing the intent of the search.

Drums · one shot · 140 BPM
PRODUCT WALKTHROUGH

See the complete workflow.

From scanning a folder to finding a focused result and moving it into a DAW—this demo shows the working desktop product end to end.

SampleFind product demoOpen on YouTube
APPLIED MULTIMODAL AI

Relevance is a system, not a single model call.

SampleFind combines fast, explainable signals with semantic audio understanding. Exact metadata stays precise; broader creative language can reach beyond the filename.

QUERY PIPELINELOCAL
“grand piano melody, 92 BPM”
TEXT + METADATAExact signalsfilename · key · BPM · type
+
AUDIO EMBEDDINGSemantic meaningtimbre · instrument · character
QUERY-AWARE HYBRID RANKING
Grand_Piano_Melody_Cm_92.wavFocused result · loop · 92 BPM · C minor01
A

Deterministic signals

Exact filenames, aliases, sample type, key, BPM, categories, and user corrections establish high-confidence relevance.

B

Audio–text intelligence

LAION-CLAP embeddings connect natural-language descriptions to what an audio sample actually sounds like.

C

Query-aware ranking

Specific instrument queries stay precise while broad regional or musical searches intentionally widen the result set.

D

Human-in-the-loop control

Manual reclassification stays private to the user and improves their library without silently retraining global behavior.

ENGINEERED END TO END

From audio analysis to desktop distribution.

SampleFind is a full-stack product spanning background processing, retrieval quality, responsive interfaces, authentication, local filesystem integration, and cross-platform release engineering.

Background scan orchestrationQueued folders, persistent progress, cancellation, rollback, and publish-on-completion.
Local data architectureSQLite metadata and user state paired with ChromaDB vector retrieval.
Producer-first desktop workflowInstant preview, layered filters, source management, and file drag-and-drop.
Cross-platform packagingVersioned macOS DMG and Windows NSIS installers delivered through GitHub Releases.
TECHNOLOGY STACK

PORTFOLIO BETA · V0.1.6

Explore the working desktop product.

Download the current build for portfolio review and controlled testing, or visit the public project repository for the complete engineering overview.