Build the Next Wave of AI With TiDB

class Chunk(TableModel):
    __tablename__ = "chunks"

    id: int = Field(primary_key=True)
    text: str = Field()
    text_vec: list[float] = text_embed.VectorField(
        source_field="text",
    )
    meta: dict = Field(sa_type=JSON)

table = db.create_table(schema=Chunk, if_exists="overwrite")

results = (
    table.search(query_text)
    .debug(True)
    .filter({"meta.language": language})
    .distance_threshold(distance_threshold)
    .limit(query_limit)
    .to_list()
)
                                                

class Pet(TableModel):
    __tablename__ = "pets"

    id: int = Field(primary_key=True)
    breed: str = Field()
    image_uri: str = Field()
    image_name: str = Field()
    image_vec: Optional[List[float]] = embed_fn.VectorField(
        distance_metric=DistanceMetric.L2,
        source_field="image_uri",
        source_type="image",
    )

results = (
    table.search(query="fluffy orange cat")
    .distance_metric(DistanceMetric.L2)
    .limit(limit)
    .to_list()
)
                                                

# Define table schema
class Document(TableModel):
    __tablename__ = "documents"

    id: int = Field(primary_key=True)
    text: str = FullTextField()
    text_vec: list[float] = embed_fn.VectorField(
        source_field="text",
    )
    meta: dict = Field(sa_type=JSON)

query = (
    table.search(query_text, search_type="hybrid")
    .distance_threshold(0.8)
    .fusion(method="rrf")
    .limit(limit)
)
                                                

# Define embedding function
embed_func = EmbeddingFunction(
    model_name="tidbcloud_free/amazon/titan-embed-text-v2"
    # No API key required for TiDB Cloud free models
)

class Chunk(TableModel):
    __tablename__ = "chunks"

    id: int = Field(primary_key=True)
    text: str = Field(sa_type=TEXT)
    text_vec: list[float] = embed_func.VectorField(source_field="text")

table = db.create_table(schema=Chunk, if_exists="overwrite")

# Insert sample data - embeddings generated automatically
table.bulk_insert([
    Chunk(text="TiDB is a distributed database that supports OLTP, OLAP, HTAP and AI workloads."),
    Chunk(text="PyTiDB is a Python library for developers to connect to TiDB."),
    Chunk(text="LlamaIndex is a Python library for building AI-powered applications."),
])
                                                

class Product(TableModel):
    __tablename__ = "products"

    id: int = Field(primary_key=True)
    name: str = Field(sa_type=TEXT)
    description: str = Field(sa_type=TEXT)
    description_vec: list[float] = embed_func.VectorField(
        source_field="description"
    )
    category: str = Field(sa_type=TEXT)
    price: float = Field()

table = db.create_table(schema=Product, if_exists="overwrite")

# App-1 is inserting strings and vectors are auto-generated in real-time
table.insert(Product(
    name="Professional Basketball",
    description="High-quality basketball for professional and amateur players",
    category="Sports",
    price=29.99
))

# App-2 is searching at once using semantic similarity with user preferences
recommendations = (
    table.search(user_profile)
    .distance_threshold(distance_threshold)
    .limit(5)
    .to_list()
)