Simpleflo
Utilities for calm, repeatable AI work -- so you can stay in flow.
Most people use AI through a narrow interface: a chat box, too many tabs, and sometimes a coding copilot. That's fine for quick answers. It breaks when the work has real context: multiple sources, competing constraints, and decisions you need to defend.
Simpleflo turns that gap into a small set of utilities with simple surfaces, careful defaults, and repeatable workflows. Use them today: decision support, grounded retrieval, and private context -- without the usual wiring tax.
The utilities
Each utility focuses on a specific kind of clarity, with workflows that stay calm under real constraints.
Decision support
Kinetiq
Decision support, not just search. Choosing AI utilities often feels like a gamble: loud claims, scattered reviews, and best-of lists that do not match your constraints. Kinetiq replaces search-and-hope with structured discovery. It uses an AI Matchmaker to interpret your needs, then recommends best-fit options while showing trade-offs, alternatives, and why each match is suggested.
Grounded retrieval
Scientia
A research assistant for experimenting with knowledge discovery and augmentation. Make RAG visible and build conviction with evidence. When retrieval-augmented answers fail, they usually fail silently. You get a confident response, and no clear way to tell whether it is grounded in your material or just plausible. Scientia helps you inspect grounding, compare retrieval setups, and explore graph-based retrieval paths so you can build an evaluation habit based on evidence, not vibes.
Private knowledge base
Conduit
Private context, safely and with less wiring. The most useful AI work is personal: your notes, your projects, your internal docs, your real workflows. Conduit aims to make private context usable, auditable, and local-first by default, while reducing the friction of connecting AI clients to the systems you already use.
Why this exists
In my professional work, I build internal systems where the constraints are high, but the artifacts are private. Simpleflo is the public demonstration of how I approach building: how I frame the problem, choose trade-offs, and handle the unglamorous details that make software trustworthy.
It reflects the standards I value:
- Clarity: Solve the user's problem, not just ship features.
- Security: Protect boundaries and secrets by default.
- Safety: Prefer grounded, verifiable outputs over confident guesses.
- Craft: Make the defaults feel obvious and keep complexity earned.
These utilities are free for the community. If they help you make a better decision or understand your own information more clearly, they have done their job.
Read the principlesLatest Insights
View all articles →Asking better questions from your knowledge
Why simple tools often outperform complex suites, and how we apply this philosophy to everything we build at Simpleflo.
Bringing private context to AI safely
AI feels magical when it has the right context — and awkward when it doesn't. The trick is that more context isn't the same thing as better context, and the difference is where usefulness and safety live.
Choosing AI utilities without guesswork
How to cut through the noise and select tools that actually fit your workflow, using the 'jobs to be done' framework.