Voice Experience
Turn Voice-AI into action
Voice Experience is the key to transform Voice-AI technologies into actionnable solutions in complex industrial contexts.
About Voice Experience
For the first time, Voice Experience® considers at the same level of importance the three essential components in the development of voice-AI interactions and intelligent conversational agents for the industry: human factors, industrial performance constraints, and software integration of voice technologies in business software.
- User acceptance criteria and expectations for an AI-based system are essential human factors used as input for specifying voice interactions and managing human-system dialogues.
- The expected performance constraints of the voice solution based on noise level, suitable audio equipment, and the working environment are used to adjust the parameters of the implemented voice assistance solution.
- The list of software already deployed, on the procution, maintenance or inspection processes (MES, CMMS, FSM, …), and on the meetings reports or projects reports (CRM, ERP, …) needs to be considered to fine the best integration method for Voice-AI interactions.
- These input data enable to optimize the configuration parameters of the SPIX Voice Assistance function, depending on the need. Its interface with business software maximizes the impact of the use of voice for manufacturers.
SPIX industry’s ambition for Voice Experience® is to constitute the first integrated platform where end-users’ wishes, and expected performance will be used as input parameters to specify and integrate Voice-AI solutions operational for the industry.
The Voice Experience proposal
The challenge to solve is to give the industrials the knowhow and the confidence for them to massively deploy voice assistance technologies for any industrial processes, from shop floor and for inspection, to meeting rooms. Nevertheless, the deployment of voice-first solutions in the industry requires the resolution of a multidimensional system of constraints: coherence of the semantic, performance in the noise, data privacy, user acceptability, audio hardware, utility and usability, and the interface to the already deployed software.
The Voice Experience manages the necessary actions required to prototype, develop and and deploy a SPIX Voice-AI assistant in the industry.
- Consider and reply to all the expectations of the workers (field operators, project leaders, inspectors, …), to the constraints of their working environment, and to their specific IT concerns.
- Reply to the fears and fantasy from the operators regarding Voice-AI technologies.
- Use a coherent and structured approach to collect user’s expectations and turn them into input data for end to end specification of Voice-AI functions.
- Based on its own tools, SPIX industry is able to analyze your soundtrack (voice + environment noise) to assess the performance of voice recognition.
- Voice-flows are tested and validated in an anechoic chamber to reproduce the noise of the working environments.
- The performance of audio hardware is validated as function of the working environment.
- The expected performance, with validated metrics, is used as input data for the end to end specification of Voice Assistance functions.
- The SPIX Dialoger is the Conversational Agent (or dialog activities manager) used to create advanced voice assistance functions for any software or complex system.
- The SPIX.Skills are off the self, embedded, robust to noise and operational offline, business Voice Services adapted to the users’ objectives and ready to be interfaces with the business software from the industry (MES, CMMS and FSM). The SPIX.Skills are build on top of the SPIX Dialoger.
- Sovereign ASR (Automatic Voice Recognition) is necessary to master the quality of the input for the SPIX Voice Assistant. The ASR from SPIX benefits from the latest development of AI technologies (SLM, LLM, embeddings, …) and is fine tuned to be operational in any industrial environment.
