AREA OF WORK

Artificial Intelligence

Unexpectech's AI work starts from a simple position: intelligent systems should adapt to people, not require people to adapt to them. The projects in this area are connected investigations into software that can evolve, agents that can persist beyond a single interface, and AI that works with human context instead of replacing human judgment.

Self-evolving softwareIntelligent agentsPerson-centered AICreative AI

Adaptive systems

Most software is frozen the moment it ships. Project Semilla asks what happens when software can participate in its own development — inspecting requirements, modifying its own source code, compiling the result, and learning from the feedback, while the person using it stays in control of what changes.

Persistent intelligent agents

An intelligent agent should not disappear when the chat window closes. IAm explores whether one agent can preserve its identity and context while manifesting through a physical robot, a virtual environment, or a conversational interface.

Human-centered AI

rAIn applies the same principle to personal context: AI assistance built around memory the person owns, permissions they control, and explanations they can inspect. Across all three projects, the goal is AI that supports human decisions and creativity — with ownership, authorization, and understandable behavior as requirements, not features.

Projects in Artificial Intelligence

Project Semilla

Active Research

A self-evolving software platform that can inspect, modify, test, and improve source code while preserving user ownership and control.

C#RoslynLocal LLMsSelf-modification

Project page

EducationAI

IAm - Intelligent Agent Manifested

Working Prototype

A persistent-agent research initiative spanning physical robotics, virtual environments, and conversational interfaces. PiDog Lab is the physical experimentation environment inside IAm.

RoboticsAgentsEmbodied AIEducation

Project page

EducationMedicineAI

rAIn

Active Research

Person-centered AI around persistent personal context, authorized support networks, semantic memory, learning context, health context, explainability, and data sovereignty.

Persistent contextExplainabilityData sovereigntySupport network

Project page

ArtsAI