Mitch Haile has extensive experience designing software for data systems in corporate environments. He was an early engineer at ground-breaking compression company Data Domain. Mitch later co-founded virtual machine backup company Pancetera, which the team sold to Quantum, a respected data backup company. Mitch has served as an executive other start-ups involved with software security, storage, and backup.
In this conversation, Mitch talks about Bit Lasso Reveal, his new program that helps search slides.
Geetesh: Mitch, can you tell us about Bit Lasso Reveal? What motivated you to create this product?
Mitch: I’ve spent a large chunk of my life building, editing, reviewing, and presenting PowerPoint and Keynote decks. I’ve always been frustrated at how hard it is to find one slide; I tend to remember slides visually. When I search my Mac with the built-in tools, I get a list of PowerPoint slides that match the queries—but the results show the first slide of the deck, not the match.
The other challenge I’ve encountered is the data overload: I may have 20 or 200 decks that contain variations of the same slide, and so I wanted a tool that can help reduce those results into something more manageable.
My background is software development, so it was “obvious” what to do next: build a solution for the slide search problem.
Geetesh: Can you share some feedback from end users of Reveal? Give us an idea of who uses Reveal, how they use it, and how their feedback has helped you.
Mitch: The Reveal customer base currently breaks down into two large categories: Professors and executives; marketers and consultants use Reveal too, but in smaller numbers. Some customers have 100,000 slides on their Macs—far more than I have, so there have been a lot of performance improvements to the product to help keep things zippy.
Across all users, I believe the use case is about the same: the customer is working on a new slide deck and knows she’s made a certain slide before, but the question is where is that slide. Presentation documents tend to be “living documents”. For example, after giving a presentation, I often want to slightly tweak things based on questions or my own perception of the flow. It doesn’t take long to find yourself cluttered with many versions of the presentations, either edited for different time slots, audiences, or our own evolution of the story. We may have master slide themes, but often a master deck eludes us.
One challenge in creating a product like Reveal is that the technology operates based on how the customer’s data is “shaped”. Do you have a lot of images, text, how are slide decks spread out across your computer or external drives. My own way of working is just one example, so customer feedback has been (and will continue to be) critical for understanding the different ways that people think about search and think about presentations.
Reveal 2.0 has a new interface that leverages your vision system in the brain more efficiently by using the whole window to organize search results visually. The idea has two parts:
- That slide “search” is usually slide “recall”. For example, a customer needs to find a specific competition slide with the green bar chart she made last year, and
- That the brain can pick our key visual characteristics subconsciously so that when the green bar chart shows up, you snap to it immediately, vs scrolling through a list manually.
Of course, with a proliferation of versions, you will probably need to narrow down the group some, and so Reveal lets you pick a “similarity group” and zoom into just that group rapidly.
There are a lot of other user-inspired features in the product as well: searching by a certain folder, showing a date histogram, opening multiple search windows at the same time, filtering search results by selecting a date range, and so on.
Reveal has a number of technologies under the hood to help deal with slides with a lot of images and new image-related features that I’m working on for future versions, which have come about from folks who are creating more modern slide decks with fewer words.