MIHR
AI Transparency Manifest
| Manifest No. | MR - AM - INF - 026 - 00002 - R0 | Issue date | 8 September 2026 |
|---|---|---|---|
| Status | Active | Expiry date | 8 September 2027 |
This document serves as a voluntary statement on the use of Artificial Intelligence (AI). It is designed to provide a structured overview of where and how AI technologies are integrated, aiming for clarity and transparency.
1. Declarant Profile
| Legal name | MIHR AI LABELING SYSTEM | Registration No. | 56627777 |
|---|---|---|---|
| Country | Armenia | Legal form | LLC |
| Licenses & certifications | |||
| Industry | IT Consulting & Services, Application Software, Cloud Services & SaaS | Sub-industry | IT Consulting, Software as a Service, Platform as a Service |
| Contact person | Larisa Ginosyan | Role / title | CEO |
| info@themihr.ai | Phone | +37493937007 | |
| Registered address | 16 Koghbatsi str, apt 63, 0010 Yerevan, Armenia | Website / public URL | https://www.themihr.ai |
| Signatory name | Larisa Ginosyan | Position / title | CEO |
2. AI Usage Overview
AI Usage Summary
The MIHR AI LABELING SYSTEM has adopted AI across 4 business domains encompassing 9 business processes, with AI-enabled activity currently comprising 43.20% of total operations. While the majority of processes — 56.80% — remain non-AI.
3. AI Application Details
The table below lists all business domains and processes in which AI is deployed, the functions performed, the technology used, and the assigned level of autonomy. The pie chart illustrates the relative weight of each process within its domain.
| Business Domain 1 | |||||
|---|---|---|---|---|---|
| Domain name | Research and Development | Stake in overall activity | 30% | ||
| Process name | Stake within domain | AI function performed | Technology used | Level of autonomy | AI/Human Roles |
| Research | 20% | Data Collection, Market Data Analysis | Claude, ChatGPT, Google Gemini, Claude — Action Items + Key Facts + Outcome | L2 - Advisory | AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions.; AI analyses data and provides structured recommendations or draft outputs. A human reviews, approves, and executes every action. |
| Application Development | 30% | AI Code Completion, Generation & Pair Programming, AI-Automated Code Review & Technical Debt Detection | Gemini Code, Google Antigravity, Composer, Automated Code Review (Diff + LLM) | L2 - Advisory | AI analyses data and provides structured recommendations or draft outputs. A human reviews, approves, and executes every action.; AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions. |
| UX/UI Design | 10% | Graphical Design | Google Gemini, ChatGPT, Claude Design | L3 - Semi-Autonomy | AI executes tasks autonomously within a defined rule-set or confidence threshold. Exceptions, edge-cases, or low-confidence outputs are escalated to a human. |
Key observations
Within the Research and Development domain, AI has been adopted across 3 business processes, representing 60% of total domain activity. The remaining 40% of processes operate without AI integration. UX/UI Design stands out as the highest-autonomy process within this domain.
| Business Domain 2 | |||||
|---|---|---|---|---|---|
| Domain name | Marketing | Stake in overall activity | 30% | ||
| Process name | Stake within domain | AI function performed | Technology used | Level of autonomy | AI/Human Roles |
| Market Research | 30% | Data Collection, Market Data Analysis, Marketing Materials Preparation | Claude, Google Gemini, ChatGPT | L3 - Semi-Autonomy | AI analyses data and provides structured recommendations or draft outputs. A human reviews, approves, and executes every action.; AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions.; AI executes tasks autonomously within a defined rule-set or confidence threshold. Exceptions, edge-cases, or low-confidence outputs are escalated to a human. |
| Digital & Social Media Communications | 5% | Social Media Content Generation | Claude, ChatGPT | L3 - Semi-Autonomy | AI executes tasks autonomously within a defined rule-set or confidence threshold. Exceptions, edge-cases, or low-confidence outputs are escalated to a human. |
| Customer Acquisition | 7% | Customer Segmentation, Customer Screening, Personalized Communications | Google Gemini, Claude, ChatGPT | L3 - Semi-Autonomy | AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions.; AI analyses data and provides structured recommendations or draft outputs. A human reviews, approves, and executes every action.; AI executes tasks autonomously within a defined rule-set or confidence threshold. Exceptions, edge-cases, or low-confidence outputs are escalated to a human. |
Key observations
Within the Marketing domain, AI has been adopted across 3 business processes, representing 42% of total domain activity. The remaining 58% of processes operate without AI integration. Market Research stands out as the highest-autonomy process within this domain.
| Business Domain 3 | |||||
|---|---|---|---|---|---|
| Domain name | Financial Planning & Analysis (FP&A) | Stake in overall activity | 12% | ||
| Process name | Stake within domain | AI function performed | Technology used | Level of autonomy | AI/Human Roles |
| Financial Forecasting & Planning | 30% | Financial Plan Creation | Claude | L1 - Assistance | AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions. |
Key observations
Within the Financial Planning & Analysis (FP&A) domain, AI has been adopted across 1 business process, representing 30% of total domain activity. The remaining 70% of processes operate without AI integration. Financial Forecasting & Planning stands out as the highest-autonomy process within this domain.
| Business Domain 4 | |||||
|---|---|---|---|---|---|
| Domain name | Legal & Corporate Secretarial | Stake in overall activity | 15% | ||
| Process name | Stake within domain | AI function performed | Technology used | Level of autonomy | AI/Human Roles |
| Contract Management & Legal Drafting | 30% | Document Drafting | Claude | L1 - Assistance | AI surfaces relevant information, suggestions, or predictions to support a human who makes all decisions and takes all actions. |
| Research | 30% | Research, AI Monitoring of Regulatory Changes | Google Gemini, Claude | L4 - High-Autonomy | AI analyses data and provides structured recommendations or draft outputs. A human reviews, approves, and executes every action.; AI operates end-to-end with minimal human touchpoints. Humans receive summary reports and retain override capability but are not part of routine execution. |
Key observations
Within the Legal & Corporate Secretarial domain, AI has been adopted across 2 business processes, representing 60% of total domain activity. The remaining 40% of processes operate without AI integration. Research stands out as the highest-autonomy process within this domain.
Disclaimer: This information is voluntarily provided by the declarant. This Manifest sets out the declarant's best efforts to describe the use of AI in order to facilitate a better understanding of the implementation of AI.
AI Transparency Manifest issued by MIHR - Machine Intelligence Human Ratio | www.themihr.ai | info@themihr.ai