Evolving our calendar assistant Reclaim to be AI-native without starting over (opens on the source site)
How we redesigned Reclaim for AI and natural-language requests while preserving the scheduling experience people rely on.
Read at the sourceTHE COMPANY INDEX TRACKED BLOG
Dropbox
Ideas, decisions, and lessons from the team.
blogs.dropbox.com (opens on the source site)LinkedIn X9Posts tracked
2 weeks agoLatest publication
0.8Posts / month over the last 12 months
Latest writing
9 of 9 postsWhat Dropbox has learned from deploying AI at company scale (opens on the source site)
Learnings from deploying AI at company scale, and what organizations need to consider as they move from AI adoption to broader transformation.
Read at the sourceTesting cookie behavior across hundreds of web surfaces with our in-house auditor (opens on the source site)
Our cookie auditor acts like a privacy-conscious user by visiting web pages and checking that they only load cookies consistent with that user’s preferences.
Read at the sourceHow our universal content processing platform Riviera evolved for AI and beyond (opens on the source site)
Riviera is the Dropbox content processing platform that’s been iteratively improving content transformation in our products for roughly a decade.
Read at the sourceHow we used DSPy to turn AI evaluations into better responses in Dash chat (opens on the source site)
We used DSPy to improve LLM judges and optimize our chat experience, creating an evaluation-driven feedback loop that produced better outputs.
Read at the sourceHow Dropbox uses MCP and Dash to close the design-to-code security gap (opens on the source site)
Using an agentic AI system to surface threat models during code review and spot gaps between security requirements and implementation.
Read at the sourceBeyond code generation: rethinking engineering productivity in the age of AI agents (opens on the source site)
How Dropbox is moving from AI tools that assist engineers to agentic systems that can execute scoped tasks, and how we’re building platforms to support those workflows.
Read at the sourceImproving storage efficiency in Magic Pocket, our immutable blob store (opens on the source site)
By turning compaction into a layered, adaptive pipeline and strengthening our monitoring and controls, we made Magic Pocket more resilient to workload changes.
Read at the sourceHow we optimized Dash's relevance judge with DSPy (opens on the source site)
We used DSPy to turn prompt engineering for our relevance judge into a measurable, automated optimization loop, improving task performance, cost, and how reliably it works in production.
Read at the source