diff --git a/backend/app/alembic/versions/043_add_project_org_to_job_table.py b/backend/app/alembic/versions/043_add_project_org_to_job_table.py
new file mode 100644
index 000000000..3bb2ccfff
--- /dev/null
+++ b/backend/app/alembic/versions/043_add_project_org_to_job_table.py
@@ -0,0 +1,92 @@
+"""Add project_id and organization_id to job table
+
+Revision ID: 043
+Revises: 042
+Create Date: 2026-02-04 14:39:00.000000
+
+"""
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "043"
+down_revision = "042"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    # Add organization_id column
+    op.add_column(
+        "job",
+        sa.Column(
+            "organization_id",
+            sa.Integer(),
+            nullable=False,
+            comment="Reference to the organization",
+        ),
+    )
+    
+    # Add project_id column
+    op.add_column(
+        "job",
+        sa.Column(
+            "project_id",
+            sa.Integer(),
+            nullable=False,
+            comment="Reference to the project",
+        ),
+    )
+    
+    # Create foreign key constraints
+    op.create_foreign_key(
+        "fk_job_organization_id",
+        "job",
+        "organization",
+        ["organization_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+    
+    op.create_foreign_key(
+        "fk_job_project_id",
+        "job",
+        "project",
+        ["project_id"],
+        ["id"],
+        ondelete="CASCADE",
+    )
+    
+    # Create indexes
+    op.create_index(
+        "ix_job_organization_id",
+        "job",
+        ["organization_id"],
+        unique=False,
+    )
+    
+    op.create_index(
+        "ix_job_project_id",
+        "job",
+        ["project_id"],
+        unique=False,
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    # Drop indexes
+    op.drop_index("ix_job_project_id", table_name="job")
+    op.drop_index("ix_job_organization_id", table_name="job")
+    
+    # Drop foreign key constraints
+    op.drop_constraint("fk_job_project_id", "job", type_="foreignkey")
+    op.drop_constraint("fk_job_organization_id", "job", type_="foreignkey")
+    
+    # Drop columns
+    op.drop_column("job", "project_id")
+    op.drop_column("job", "organization_id")
+    # ### end Alembic commands ###
diff --git a/backend/app/alembic/versions/044_optimize_conversation_query.py b/backend/app/alembic/versions/044_optimize_conversation_query.py
new file mode 100644
index 000000000..d08e68cf5
--- /dev/null
+++ b/backend/app/alembic/versions/044_optimize_conversation_query.py
@@ -0,0 +1,34 @@
+"""add composite index for conversation query optimization
+
+Revision ID: 044
+Revises: 043
+Create Date: 2026-02-04 15:24:00.000000
+
+"""
+from alembic import op
+
+
+# revision identifiers, used by Alembic.
+revision = "044"
+down_revision = "043"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # Create composite index to optimize the get_conversation_by_ancestor_id query
+    # This query filters by: ancestor_response_id, project_id, is_deleted
+    # and orders by: inserted_at DESC
+    op.create_index(
+        "ix_openai_conversation_ancestor_project_active_time",
+        "openai_conversation",
+        ["ancestor_response_id", "project_id", "is_deleted", "inserted_at"],
+        unique=False,
+    )
+
+
+def downgrade():
+    op.drop_index(
+        "ix_openai_conversation_ancestor_project_active_time",
+        table_name="openai_conversation",
+    )
diff --git a/backend/app/api/docs/documents/upload.md b/backend/app/api/docs/documents/upload.md
index e667015f5..c6e51bbe3 100644
--- a/backend/app/api/docs/documents/upload.md
+++ b/backend/app/api/docs/documents/upload.md
@@ -4,6 +4,12 @@ Upload a document to Kaapi.
 - If a target format is specified, a transformation job will also be created to transform document into target format in the background. The response will include both the uploaded document details and information about the transformation job.
 - If a callback URL is provided, you will receive a notification at that URL once the document transformation job is completed.
 
+### File Size Restrictions
+
+- **Maximum file size**: 50MB (configurable via `MAX_DOCUMENT_UPLOAD_SIZE_MB` environment variable)
+- Files exceeding the size limit will be rejected with a 413 (Payload Too Large) error
+- Empty files will be rejected with a 422 (Unprocessable Entity) error
+
 ### Supported Transformations
 
 The following (source_format → target_format) transformations are supported:
diff --git a/backend/app/api/routes/documents.py b/backend/app/api/routes/documents.py
index 58beb31b8..3e42e9952 100644
--- a/backend/app/api/routes/documents.py
+++ b/backend/app/api/routes/documents.py
@@ -34,6 +34,7 @@
     build_document_schema,
     build_document_schemas,
 )
+from app.services.documents.validators import validate_document_file
 from app.utils import (
     APIResponse,
     get_openai_client,
@@ -123,6 +124,9 @@ async def upload_doc(
     if callback_url:
         validate_callback_url(callback_url)
 
+    # Validate file size before uploading to S3
+    await validate_document_file(src)
+
     source_format, actual_transformer = pre_transform_validation(
         src_filename=src.filename,
         target_format=target_format,
diff --git a/backend/app/core/config.py b/backend/app/core/config.py
index 40c770541..eaab6a6e0 100644
--- a/backend/app/core/config.py
+++ b/backend/app/core/config.py
@@ -123,6 +123,9 @@ def AWS_S3_BUCKET(self) -> str:
     CALLBACK_CONNECT_TIMEOUT: int = 3
     CALLBACK_READ_TIMEOUT: int = 10
 
+    # Document upload size limit (in MB)
+    MAX_DOCUMENT_UPLOAD_SIZE_MB: int = 512
+
     @computed_field  # type: ignore[prop-decorator]
     @property
     def COMPUTED_CELERY_WORKER_CONCURRENCY(self) -> int:
diff --git a/backend/app/crud/jobs.py b/backend/app/crud/jobs.py
index 1a9005b69..f6da6b6f0 100644
--- a/backend/app/crud/jobs.py
+++ b/backend/app/crud/jobs.py
@@ -12,9 +12,17 @@ class JobCrud:
     def __init__(self, session: Session):
         self.session = session
 
-    def create(self, job_type: JobType, trace_id: str | None = None) -> Job:
+    def create(
+        self,
+        job_type: JobType,
+        project_id: int,
+        organization_id: int,
+        trace_id: str | None = None,
+    ) -> Job:
         new_job = Job(
             job_type=job_type,
+            project_id=project_id,
+            organization_id=organization_id,
             trace_id=trace_id,
         )
         self.session.add(new_job)
diff --git a/backend/app/models/job.py b/backend/app/models/job.py
index b6a1a5ae7..753d1ab33 100644
--- a/backend/app/models/job.py
+++ b/backend/app/models/job.py
@@ -1,11 +1,16 @@
 from datetime import datetime
 from enum import Enum
+from typing import TYPE_CHECKING, Optional
 from uuid import UUID, uuid4
 
-from sqlmodel import Field, SQLModel
+from sqlmodel import Field, Relationship, SQLModel
 
 from app.core.util import now
 
+if TYPE_CHECKING:
+    from .organization import Organization
+    from .project import Project
+
 
 class JobStatus(str, Enum):
     PENDING = "PENDING"
@@ -58,6 +63,22 @@ class Job(SQLModel, table=True):
         },
     )
 
+    # Foreign keys
+    organization_id: int = Field(
+        foreign_key="organization.id",
+        nullable=False,
+        ondelete="CASCADE",
+        index=True,
+        sa_column_kwargs={"comment": "Reference to the organization"},
+    )
+    project_id: int = Field(
+        foreign_key="project.id",
+        nullable=False,
+        ondelete="CASCADE",
+        index=True,
+        sa_column_kwargs={"comment": "Reference to the project"},
+    )
+
     # Timestamps
     created_at: datetime = Field(
         default_factory=now,
@@ -68,6 +89,10 @@ class Job(SQLModel, table=True):
         sa_column_kwargs={"comment": "Timestamp when the job was last updated"},
     )
 
+    # Relationships
+    organization: Optional["Organization"] = Relationship()
+    project: Optional["Project"] = Relationship()
+
 
 class JobUpdate(SQLModel):
     status: JobStatus | None = None
diff --git a/backend/app/services/documents/validators.py b/backend/app/services/documents/validators.py
new file mode 100644
index 000000000..8f5fcfba6
--- /dev/null
+++ b/backend/app/services/documents/validators.py
@@ -0,0 +1,49 @@
+"""Validation utilities for document uploads."""
+
+import logging
+from pathlib import Path
+
+from fastapi import HTTPException, UploadFile
+
+from app.core.config import settings
+from app.utils import mask_string
+
+logger = logging.getLogger(__name__)
+
+# Maximum file size for document uploads (in bytes)
+# Default: 512 MB, configurable via settings
+MAX_DOCUMENT_SIZE = settings.MAX_DOCUMENT_UPLOAD_SIZE_MB * 1024 * 1024
+
+
+async def validate_document_file(file: UploadFile) -> int:
+    """
+    Validate document file size.
+
+    Args:
+        file: The uploaded file
+
+    Returns:
+        File size in bytes if valid
+    """
+
+    # Get file size by seeking to end
+    file.file.seek(0, 2)
+    file_size = file.file.tell()
+    file.file.seek(0)
+
+    if file_size > MAX_DOCUMENT_SIZE:
+        raise HTTPException(
+            status_code=413,
+            detail=f"File too large. Maximum size: {MAX_DOCUMENT_SIZE / (1024 * 1024):.0f}MB",
+        )
+
+    if file_size == 0:
+        raise HTTPException(
+            status_code=422,
+            detail="Empty file uploaded"
+        )
+
+    logger.info(
+        f"[validate_document_file] Document file validated: {mask_string(file.filename)} ({file_size} bytes)"
+    )
+    return file_size
diff --git a/backend/app/services/llm/jobs.py b/backend/app/services/llm/jobs.py
index f4700b51b..793c5bf2e 100644
--- a/backend/app/services/llm/jobs.py
+++ b/backend/app/services/llm/jobs.py
@@ -26,7 +26,12 @@ def start_job(
     """Create an LLM job and schedule Celery task."""
     trace_id = correlation_id.get() or "N/A"
     job_crud = JobCrud(session=db)
-    job = job_crud.create(job_type=JobType.LLM_API, trace_id=trace_id)
+    job = job_crud.create(
+        job_type=JobType.LLM_API,
+        project_id=project_id,
+        organization_id=organization_id,
+        trace_id=trace_id,
+    )
 
     try:
         task_id = start_high_priority_job(
diff --git a/backend/app/services/response/jobs.py b/backend/app/services/response/jobs.py
index cab0f0e83..2b609b77d 100644
--- a/backend/app/services/response/jobs.py
+++ b/backend/app/services/response/jobs.py
@@ -19,7 +19,12 @@ def start_job(
     """Create a response job and schedule Celery task."""
     trace_id = correlation_id.get() or "N/A"
     job_crud = JobCrud(session=db)
-    job = job_crud.create(job_type=JobType.RESPONSE, trace_id=trace_id)
+    job = job_crud.create(
+        job_type=JobType.RESPONSE,
+        project_id=project_id,
+        organization_id=organization_id,
+        trace_id=trace_id,
+    )
 
     try:
         task_id = start_high_priority_job(
diff --git a/backend/app/tests/api/routes/documents/test_route_document_upload.py b/backend/app/tests/api/routes/documents/test_route_document_upload.py
index 6f16b52b1..3a4c78040 100644
--- a/backend/app/tests/api/routes/documents/test_route_document_upload.py
+++ b/backend/app/tests/api/routes/documents/test_route_document_upload.py
@@ -325,3 +325,102 @@ def test_upload_response_structure_without_transformation(
             assert field in response.data
 
         assert response.data["transformation_job"] is None
+
+    @patch("app.services.documents.validators.MAX_DOCUMENT_SIZE", 1024 * 1024)  # Mock to 1MB
+    def test_upload_file_exceeds_size_limit(
+        self,
+        db: Session,
+        route: Route,
+        uploader: WebUploader,
+    ) -> None:
+        """Test that files exceeding the size limit are rejected."""
+        aws = AmazonCloudStorageClient()
+        aws.create()
+
+        # Create a file larger than the 1MB mock limit
+        # For testing purposes, we'll create a 2MB file
+        with NamedTemporaryFile(mode="wb", suffix=".pdf", delete=False) as fp:
+            # Write 2MB of data (2 * 1024 * 1024 bytes)
+            chunk_size = 1024 * 1024  # 1MB chunks
+            for _ in range(2):
+                fp.write(b"0" * chunk_size)
+            fp.flush()
+            large_file = Path(fp.name)
+
+        try:
+            response = uploader.put(route, large_file)
+            
+            assert response.status_code == 413
+            error_data = response.json()
+            assert "File too large" in error_data["error"]
+            assert "Maximum size: 1MB" in error_data["error"]
+            
+            # Verify no document was created in the database
+            statement = select(Document).where(Document.fname == str(large_file))
+            result = db.exec(statement).first()
+            assert result is None
+        finally:
+            large_file.unlink()
+
+    def test_upload_empty_file(
+        self,
+        db: Session,
+        route: Route,
+        uploader: WebUploader,
+    ) -> None:
+        """Test that empty files are rejected."""
+        aws = AmazonCloudStorageClient()
+        aws.create()
+
+        # Create an empty file
+        with NamedTemporaryFile(mode="w", suffix=".txt", delete=False) as fp:
+            # Don't write anything, just create an empty file
+            fp.flush()
+            empty_file = Path(fp.name)
+
+        try:
+            response = uploader.put(route, empty_file)
+            
+            assert response.status_code == 422
+            error_data = response.json()
+            assert "Empty file uploaded" in error_data["error"]
+            
+            # Verify no document was created in the database
+            statement = select(Document).where(Document.fname == str(empty_file))
+            result = db.exec(statement).first()
+            assert result is None
+        finally:
+            empty_file.unlink()
+
+    @patch("app.services.documents.validators.MAX_DOCUMENT_SIZE", 512 * 1024 * 1024)
+    def test_upload_file_within_size_limit(
+        self,
+        db: Session,
+        route: Route,
+        uploader: WebUploader,
+    ) -> None:
+        """Test that files within the size limit are accepted."""
+        aws = AmazonCloudStorageClient()
+        aws.create()
+
+        # Create a 1MB file (well within the limit)
+        with NamedTemporaryFile(mode="wb", suffix=".pdf", delete=False) as fp:
+            # Write 1MB of data
+            fp.write(b"0" * (1024 * 1024))
+            fp.flush()
+            normal_file = Path(fp.name)
+
+        try:
+            response = httpx_to_standard(uploader.put(route, normal_file))
+            
+            assert response.success is True
+            assert "id" in response.data
+            doc_id = response.data["id"]
+            
+            # Verify document was created in database
+            statement = select(Document).where(Document.id == doc_id)
+            result = db.exec(statement).one()
+            assert result.fname == str(normal_file)
+        finally:
+            normal_file.unlink()
+