PRODUCTS / 01—07

Start with what
you want to do.

Conversation, drafting, calls, investigation, review and publishing.
Choose a task, then explore the demonstration and requirements.

7 products

These are independently developed products. Screens are sourced from public project records; they do not guarantee current availability or production readiness.

Real app workspace
Recorded product screenFrom a public GitHub record
01 — GenieDeveloper preview

Give an idea somewhere to become work.

Turn rough notes into a plan, website copy, or a small HTML prototype. Reopen the result, copy it, or save it as Markdown.

The app needs local services and a configured model, on macOS 14 or later. The download alone is not a hosted service. A screenshot is optional and goes to the connected model only when you send a question. Previously Astra.

Evidence & reuse conditions

What you can inspect

Real app recordings and saved artifacts. Waiting is condensed.

Reuse conditions

Public source · no project-wide license

Potential project scope

Prototype an internal AI workspace or an assistant using local and external models.

Validation report
Recorded product screenFrom a public GitHub record
02 — AI MeetingBeta

Think out loud. Keep the next step.

Talk with an AI character, change direction, and keep the tasks that matter. Uncertain changes stay pending until they are confirmed.

Account-free tasks and the verified-email, time-limited voice trial are different experiences. External meeting connections require setup and evaluation.

Evidence & reuse conditions

What you can inspect

45-second recording with real Gemini responses, synthetic speech input, and some scripted confirmations.

Reuse conditions

Apache-2.0 · separate asset terms

Potential project scope

Prototype voice-based training, task capture, or an AI character experience.

Completion evidence
Recorded product screenFrom a public GitHub record
03 — OathraDeveloper release

A phone call. A result you can inspect.

A runtime for AI phone negotiations and evidence-based completion checks. Tie dates, amounts, and confirmation to the other party’s words.

A spoken agreement does not prove a booking exists in a business system. Simulation and real-call evidence are separate. Real calls need configuration and incur costs.

Evidence & reuse conditions

What you can inspect

Transcript checks, simulation (recorded calls by the built-in agent, GPT-4o mini and Gemini Flash), and documented real-call development. The 100-real-call evaluation is not complete.

Reuse conditions

Apache-2.0

Potential project scope

Prototype a phone workflow, add completion checks to a voice agent, or design consent and human handoff.

Investigation case
Recorded product screenFrom a public GitHub record
04 — AI SecureResearch prototype

From scattered alerts to one reviewable case.

Connect exposure, privileged access, and file activity. Keep observations, hypotheses, and unknowns distinct as you investigate.

Snapshot analysis, not continuous monitoring or enforcement. It does not confirm exfiltration. Published evaluations use synthetic data, not production performance guarantees.

Evidence & reuse conditions

What you can inspect

An interactive synthetic case and reproducible single-rule versus correlation evaluation.

Reuse conditions

MIT

Potential project scope

Prototype log ingestion and evidence-aware investigation. Not a replacement for production security controls.

Agent work trail
Recorded product screenFrom a public GitHub record
05 — Agent TeamUnder evaluation

Ask once. Keep only the version that was checked.

AI teammates share one request. A teammate other than the maker checks each revision, and you keep which version passed what — plus whatever was left unfinished.

A quality advantage over a single agent has not been established. Published comparisons include failures and provider constraints. Production acceptance is separate.

Evidence & reuse conditions

What you can inspect

Real-model work records and evaluation results including failures. The featured research replay ended partial.

Reuse conditions

MIT · repository: Multibot

Potential project scope

Prototype agent workflows with explicit roles, artifact checks, budgets, permissions, and human approval.

Generated launch page
Recorded product screenFrom a public GitHub record
06 — LaunchloomAlpha

You built it. Now let people see it.

Make a landscape film, a vertical cut, a landing page, and social drafts from one product brief and a real recording, locally.

Single-operator alpha. Generating material is not publishing it. Real social-account publishing remains unverified; engagement and business outcomes are not guaranteed.

Evidence & reuse conditions

What you can inspect

A film made by the tool itself, a real launch kit, and local verification records.

Reuse conditions

Apache-2.0 · separate integration terms

Potential project scope

Design a recording-to-launch workflow, brand-specific templates, and approval steps.

A scene from the introduction film (recreated screen)
Recorded product screenFrom a public GitHub record
07 — 星藍ノアStreaming

Comments, answered in voice and expression.

The reply, the voice and the look are separate systems. Noa joins them and answers stream comments with voice and expression.

It is not a system that can never give an inappropriate reply. The conversation AI has usage fees, and running it uses electricity. A recording of a full live run has not been published.

Evidence & reuse conditions

What you can inspect

A film made to explain Noa; not a recording of a live stream.

Reuse conditions

Closed source

Potential project scope

Character-led streams, guidance in shops and venues, and prototypes for handling enquiries.

START A CONVERSATION

What would you like
to make possible?

An unfinished idea is a good place to start. Tell us about the work, and what you want to try.

Work with us