Skip to main content
DEISO Mirai AI Pro v3.3.2 represents the last major version of Mirai before the v4.0 platform release in April 2026. This page documents the structured prompt framework, feature categories, and system-level behaviors that were active as of February 16, 2025. If you are reviewing version history, tracking how Mirai evolved, or comparing current capabilities against the earlier architecture, this reference captures what the system looked like at that point in time. Release date: February 16, 2025
Version: v3.3.2
Product: DEISO Mirai AI Pro
This is a legacy version reference. The features and prompt framework documented here describe Mirai as it existed before the v4.0 major release. Most capabilities listed below have been superseded or significantly enhanced in v4.0 and later versions. For current capabilities, see the v4.0, v4.1, v4.2, and v4.3 release notes.

Context: the version before v4.0

v3.3.2 was a prompt-table version — its capabilities were defined primarily through a structured prompt framework that governed how Mirai responded to different categories of user input. The system operated as a DEISO-branded AI assistant focused on blog recommendations, service matching, training comparison, country-specific guidance, and human escalation routing. v4.0, released on April 17, 2026, replaced this architecture with a unified interface, a ChatGPT 5.2 model upgrade, execution-path intelligence, structured comparison logic, and a deeper enterprise guidance layer. The sections below document what existed in v3.3.2 before those changes.

Legacy capability overview

Insights and blog recommendations

In v3.3.2, you could ask Mirai to suggest relevant blog posts from the DEISO Insights archive. The system matched your topic against a library of DEISO content covering LCA, Scope 3, circular economy, sustainability reporting, and related subjects, then returned a relevant recommendation. Example prompts from this era:
  • “Suggest a DEISO blog post on LCA”
  • “Show me a blog post about sustainability in packaging”
  • “Recommend an article from the DEISO blog about Scope 3 emissions”

Service recommendations and smart matching

Mirai v3.3.2 supported query-to-service matching across DEISO’s portfolio. You could describe a need and receive a matched service recommendation based on topic, geography, and project type. Example prompts:
  • “What DEISO service is suitable for EPD creation in France?”
  • “I’m looking for carbon footprint consulting in Japan”
  • “Do you offer training on openLCA in Saudi Arabia?”
  • “Help me choose between DEISO online and onsite sessions”

Country-specific and project-specific assistance

The system could provide localized service guidance based on country or sector context. You could describe a geography or project environment and receive DEISO service recommendations aligned to those conditions. Example prompts:
  • “What services are available for companies in Germany?”
  • “Which DEISO services apply to construction projects in UAE?”
  • “Is there a consulting service for CO₂ reporting in the US?”

Training programs, levels, and comparison

v3.3.2 included a structured training comparison capability. You could ask Mirai to compare openLCA training levels, retrieve PCF training duration, find EPD training costs, or identify which programs included certification. Example prompts:
  • “Compare DEISO openLCA training levels”
  • “Show all available PCF trainings and their duration”
  • “What’s the cost of DEISO EPD training?”
  • “Which DEISO trainings include certification?”

AI assistant and help commands

v3.3.2 included a built-in help layer that you could trigger with simple commands. This returned categorized feature guidance, example prompts, and a summary of what the system could do. Example commands:
  • “Help”
  • “What can you do?”
  • “Show me example prompts”
  • “List all features of Mirai AI Pro”

Human support and WhatsApp handoff

Before the v4.3 escalation button was introduced, v3.3.2 handled human support requests through direct prompt matching. If you asked to speak with someone or reach DEISO support, Mirai returned WhatsApp contact information and DEISO support options. Example prompts:
  • “I want to speak to someone at DEISO”
  • “Can I talk to a real expert?”
  • “How do I reach customer support?”
  • “Is there a WhatsApp contact?”

System-level output features

At the system level, v3.3.2 automatically included human support contact options in relevant responses, displayed WhatsApp access and DEISO contact information, included a version footer with messaging such as “Mirai AI Pro v3.x”, and maintained DEISO-branded AI assistant positioning throughout the earlier persona versions.

v3.3.2 feature table

#FeaturePrompt exampleDescription
1Insights-Based Blog Suggestions”Suggest a DEISO blog post on circular economy”Recommends relevant DEISO blog content based on intent.
2WhatsApp Handoff”Can I speak with someone?”Triggers human-support escalation via WhatsApp.
3Smart Service Matching”I need help with EPD for construction”Maps your query to the correct DEISO service.
4Dynamic Version FooterAutomaticDisplays synchronized system version messaging.
5Training Comparison”Compare openLCA training levels”Generates structured training comparisons.
6Help Command”Help”Returns categorized feature guidance.
7Country-Specific Guidance”Carbon reporting in Japan?”Provides localized DEISO recommendations.
8Intent Detection”Training in Saudi Arabia?”Combines service and content guidance based on detected intent.
9Branding IntegrationSystem-levelEnsures DEISO tone and structure throughout all responses.
10Chat FooterAutomaticIncludes contact information and a call-to-action in each response.
All ten features listed above were superseded or significantly enhanced in v4.0 and later versions. Blog recommendations were replaced by execution-path intelligence. WhatsApp handoff was replaced by the structured human escalation button introduced in v4.3. Smart service matching was deepened into contextual intelligence in v4.2. For current capabilities, refer to the v4.x release notes.

Historical context

v3.3.2 established the foundational prompt architecture for DEISO Mirai. It demonstrated that an AI guidance layer could support service discovery, training navigation, content recommendations, and human escalation within a single branded assistant. The v4.0 major release built directly on that foundation, replacing the prompt-table model with a more structured, model-powered, and commercially integrated execution-guidance system.