Inovativi logo
ENDE
← Case studies
Internal R&D Showcase

Roboko — AI Companion Robot Showcase

Character-driven AI companion robot concept combining voice agents, 3D-printable bodies, customization flows, simple robotics, and responsible AI framing.

View the live demo →

Context

Roboko is an internal Inovativi R&D showcase exploring how AI companions can move from pure software into physical, customizable objects. The project demonstrates a gallery of companion robot characters, a configuration flow for colors, personality, voice, accessories, and purpose, plus a technical architecture connecting speech-to-text, LLM personality agents, text-to-speech, robot controllers, LEDs, speakers, and simple actuators.

Problem

AI companions are becoming a major software category, but the physical expression of these systems is still early. The hard part is bridging a credible voice-agent experience with a physical object that can be designed, customized, manufactured, and adapted for real settings — without overclaiming humanoid capability or drifting into toy-like positioning. Roboko explores a practical middle ground: not a complex humanoid robot, but a character-driven, customizable, 3D-printable companion object that can speak, react with lights or simple motion, and be adapted for education, brand mascots, events, reception desks, and innovation showcases.

What was built / modernized

Roboko brings together a character and customization system, an AI voice-agent architecture, and a 3D-printable physical prototyping concept into one coherent showcase. Users browse a gallery of companion robot characters and move through a configuration flow that adapts colors, personality, voice, accessories, and intended purpose. Behind each character, a voice pipeline connects speech-to-text, an LLM personality agent, and text-to-speech to a robot controller that drives LEDs, speakers, and simple actuators. The experience is delivered bilingually in English and German, with responsible AI framing kept in view throughout.

AI voice-agent architecture

  1. User Voice Input

    Spoken request captured by the companion

  2. Speech-to-Text

    Transcribes audio into text

  3. LLM Personality Agent

    Character-specific persona, tone, and guardrails

  4. Text-to-Speech

    Generates the character's spoken reply

  5. Robot Controller

    Coordinates output across hardware channels

  6. LEDs / Speaker / Actuators

    Light, audio, and simple motion reactions

How a spoken request flows from microphone to a character-driven response expressed through audio, light, and simple motion.

Character & customization flow

  1. Character Gallery

    Browse companion robot characters

  2. Color & Body

    Select colors and a 3D-printable body

  3. Personality

    Choose tone and behavior of the voice agent

  4. Voice

    Pick the speaking voice

  5. Accessories

    Add modular parts and accessories

  6. Purpose

    Education, mascot, events, reception, or showcase

The configuration path that turns a base companion character into a customized, purpose-specific object.

Production judgment — what we deliberately do not allow

  • Roboko is not positioned as ready for mass production.
  • Roboko is not a certified consumer robot.
  • Roboko is not a therapeutic companion or medical device.
  • Roboko is not a fully autonomous humanoid.
  • Roboko is not a replacement for human relationships.
  • Roboko is not presented as a finished commercial robotics product.

Workflow highlights

  • Gallery of companion robot characters with distinct personalities
  • Configuration flow for colors, personality, voice, accessories, and purpose
  • Voice-agent pipeline: speech-to-text, LLM personality agent, and text-to-speech
  • Robot controller driving LEDs, speakers, and simple actuators
  • 3D-printable bodies designed for modular hardware and accessories
  • Bilingual product experience in English and German

Security, auditability & governance

  • Responsible AI framing applied to a companion-like interface, not a human substitute
  • Personality agents bounded by per-character guardrails and tone
  • Positioned as an internal prototype / showcase concept, never as a certified product
  • Capabilities described honestly — speaking, light, and simple motion, not autonomous humanoid behavior
  • Bilingual copy reviewed for consistent, non-overclaiming positioning

Value delivered

  • Voice-agent and LLM personality design
  • AI product configuration workflows
  • Responsible AI guardrails for companion-like interfaces
  • Rapid prototyping with 3D printing and modular hardware
  • Bilingual product experience in English and German
  • R&D storytelling for investors, institutions, and innovation centers
  • Practical AI + robotics architecture without overclaiming humanoid capability

Technologies

  • Voice Agents
  • Speech-to-Text
  • LLM Personality Agents
  • Text-to-Speech
  • Next.js
  • Bilingual i18n (EN / DE)
  • 3D Printing
  • Microcontrollers
  • LEDs & Actuators
  • Modular Hardware

Relevant roles

  • AI / Voice Agent Engineer
  • LLM Personality Designer
  • Product & Industrial Designer
  • 3D Printing & Hardware Prototyper
  • Bilingual Product UX Designer

Status & transparency

Internal prototype / showcase concept. Roboko is not currently positioned as a certified consumer robotics product. Its purpose is to demonstrate Inovativi's ability to connect AI engineering, product design, rapid prototyping, and advanced manufacturing workflows.

Next step

Discuss a similar project

We can adapt this pattern to your systems and provide the engineers to build it. Reach us at info@inovativi.com.

Discuss a similar projectRequest engineer profiles