The product decision
Make original sources the foundation of the product. Search finds candidates; application code retrieves exact passages and citations. Model assistance helps interpret the inquiry without becoming the authority for the text.
Knowledge & native AI
A legacy library becomes an intelligent research product.
From recovering the source database to native macOS and iOS applications, local hybrid search, and on-device AI.
How can people follow a question through a complex body of knowledge?
01
Source foundation
Recovered corpus · Stable passage identities
02
Intelligence
Local hybrid search · On-device models
03
Product experience
Native macOS & iOS · Reading and inquiry
What was built
Dimas carried nearly all product definition, database conversion, experience design, architecture, and native development. The case records describe built macOS and iOS applications with local hybrid search, including on-device Apple Foundation Models integration on iOS.
The judgment behind the architecture
A recorded evaluation rejected two semantic-search candidates when they failed the improvement criteria. The working baseline stayed active. The later hybrid implementation is a separate milestone, with no claim of benchmark superiority.
Current stage
The extended research design includes bounded tools, source-linked synthesis, evidence capture, and continuity when a reader branches and returns. Those workflows and their usability remain development and evaluation work. Adoption and commercial outcomes have not been established.
Selected experience of BuildFutures founder Dimas. Adapted from the owner’s September 2026 case records; project stage and responsibilities are stated for this engagement.
Let’s make
something
useful.
Bring an early idea, a defined brief, or a system that needs to work better.
Your first step
Shape the blueprint.
Define the problem, the first useful scope, and the decisions needed to build it.