back_office_ops · saas · workflow
Dropbox Dash uses context engineering to improve agentic AI accuracy and performance
Dash's original RAG pipeline handled basic retrieval well but was insufficient for agentic tasks; adding more tools caused analysis paralysis, context rot, and accuracy degradation on longer-running jobs.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · User submits agentic request
A user request—such as asking Dash to open the editor and write an executive summary—initiates the agentic workflow.
Tools used
DashModel Context ProtocolDash Search
Outcome
Dropbox adopted three context engineering strategies—consolidating retrieval into a single unified index, filtering context via a knowledge graph, and delegating complex query tasks to a specialized search agent—resulting in faster, more accurate agentic performance with leaner context usage.
What failed first
Using multiple separate retrieval APIs proved unreliable—the model often failed to call all required sources—and MCP tool definitions consumed excessive tokens, degrading performance.
Results
Volumeslower, less accurate decision making
Grounding & classification
Source type: technical build writeup
25 fields verified against source quotes.
agentic workflowenterprise searchknowledge searchmulti agent workflowragknowledge basefailure mode describednamed customerproduction runtime claimedtools describedworkflow describedsoftwareaccuracy improvementemployee productivitytechnical build writeupback office opsagentic task executionrag answering