Starter atlas, checked 17 August 2026
A sourced starter atlas of Armenia's AI ecosystem
Seven public entry points across research, education, infrastructure, and ecosystem support, with direct sources and visible gaps.
Armenia's AI ecosystem is not one institution or one headline. It includes public programs, research laboratories, learning networks, shared infrastructure, and organizations that connect these layers. This first atlas deliberately starts small and links every entry to a page maintained by the institution itself.
01
What we know now
- 01
The starter set uses official government, university, foundation, laboratory, and organization-owned pages.
- 02
Every entry states what the source supports and avoids inferring access or impact that the page does not document.
- 03
The current atlas is a curated starting set, not a census of Armenian AI activity.
02
Why this matters for Armenia
A useful ecosystem map helps people find a real doorway and inspect the source behind it. Inclusion here records a verifiable public function, not rank, endorsement, or complete coverage.
03
public and support
Ministry platform and EIF networkuniversity research
YSU and AUA centersindependent research
YerevaNNeducation networks
FAST programs and TUMOCounts describe this starter selection, not the size of Armenia's full ecosystem.
04
How an entry qualifies
An entry needs a current public page that identifies the organization and a concrete AI-related function in Armenia. That function may be research, education, compute, infrastructure, or ecosystem support. A broad claim that an organization is “innovative” is not enough.
Organization-owned pages are direct evidence of what the organization says it does, not independent proof of quality or impact. The atlas therefore records the supported function and the date Imananq reviewed the source without ranking the organization.
- Named organization and Armenia connection
- Concrete public AI-related function
- Direct source URL and review date
- No ranking, endorsement, or inferred performance
05
Public programs and ecosystem support
The Ministry's AI Virtual Institute describes a platform for innovation and collaboration, including the 2026 AWS compute-support pilot. FAST publishes programs across education, research, and commercialization, including Generation AI and AI learning activities. EIF lists projects spanning Engineering City, technology centers, a national supercomputing initiative, incubation, and training.
These are broad entry points. A person should follow the specific program page to confirm whether an application is open, who qualifies, what is supplied, and whether the activity is current.
- AI Virtual Institute: public program and compute-support doorway.
- FAST: education, research, and commercialization program directory.
- EIF: infrastructure, entrepreneurship, and regional technology project directory.
06
Research and learning nodes
Yerevan State University documents the YSU–Krisp AI Lab and a machine-learning research agenda linked to students and university programs. The American University of Armenia's Engineering Research Center lists interdisciplinary projects using machine learning and AI. YerevaNN describes itself as a non-profit research laboratory and publishes research, code, courses, and educational programs.
TUMO's Armenia and organization pages describe technology and design learning for teenagers, including a generative-AI track. These four nodes serve different ages and purposes; the atlas should not collapse a youth learning program, a university center, and a research lab into one category.
- YSU–Krisp AI Lab: university machine-learning research and student connection.
- AUA Engineering Research Center: interdisciplinary applied research projects.
- YerevaNN: non-profit machine-learning research and education.
- TUMO: youth technology and design learning with a generative-AI track.
07
What is missing from version one
The starter set does not yet provide a company census, regional coverage score, open job list, shared dataset directory, accelerator comparison, infrastructure benchmark, or validated map of relationships. It also does not infer collaboration merely because two organizations appear in the same ecosystem.
Expansion should happen by category and evidence rule. Candidate entries need a stable source, a clear function, and a correction path. Duplicates, stale programs, unsupported marketing claims, and profiles centred on private individuals should remain out.
- Add regions and remote-access options.
- Separate open applications from general organization pages.
- Record facilities, datasets, courses, and research groups as distinct resource types.
- Add relationship edges only when a source documents the relationship.
08
Working directory
Ministry platform with collaboration programs and the 2026 AWS compute-support pilot.
Program directory spanning science education, research support, and commercialization.
Machine-learning laboratory with a research agenda and links to university teaching and student work.
Project network including Engineering City, technology centers, incubation, and computing infrastructure initiatives.
Interdisciplinary center publishing current projects that include machine learning and AI methods.
Non-profit research laboratory publishing AI research, models, code, and educational activity.
Technology and design learning network for teenagers that includes a generative-AI track.
09
Use the atlas as a doorway
Choose the function you need, then verify the current terms at the source.
- 01
For compute or a public support program, begin with the AI Virtual Institute.
- 02
For science and education programs, inspect FAST's current program directory.
- 03
For academic research, compare the specific work of YSU and AUA centers.
- 04
For independent machine-learning research, inspect YerevaNN's publications and code.
- 05
For youth learning, review TUMO's current locations and program details.
- 06
For infrastructure and incubation, follow the relevant EIF project page.
10
Limits of this edition
This is a curated starter atlas, not an exhaustive directory or market map.
Most sources are maintained by the listed organization and do not independently establish quality or impact.
Inclusion is not endorsement, and absence does not imply that an organization is unimportant.
SRC
Source desk
Direct links to the material behind this selection. Seeing the source matters as much as reading the synthesis.