Ashot Arzumanyan

Partner at @SmartGateVC. Investing early stage in physical AI/robotics, neurotech/BCIs, cybersecurity.

[Ashot Arzumanyan @AshotVC]

@modelfront @SmartGateVC @OpenLocTMS 💥 boom :)

[Perceptron AI @perceptroninc]

Mk1.5 allows developers to build any embodied agent without training. Watch Mk1.5 control a quadruped to navigate a room, find the router in a server room, and identify the root cause of the issue.

[Ashot Arzumanyan @AshotVC]

@perceptroninc @ArmenAgha Congrats!

[Perceptron AI @perceptroninc]

Today we're releasing Mk1.5: a new intelligence layer for embodied agents.

It flies drones, controls quadrupeds, powers smart glasses, tracks objects, searches the web, reasons visually, and dispatches its own sub-agents.

One model, no platform-specific retraining. 🧵

[EVN Report @evn_report]

As AI reshapes the global technology industry, Armenia’s traditional advantage in software development is under threat. Ashot Arzumanyan argues that the country must reinvent its tech ecosystem around physical AI to remain competitive.
https://t.co/gtamT5uvHp

[Perceptron AI @perceptroninc]

Today we're releasing Isaac 0.5: 36B dynamic MoE, open weight 🤗 embodied foundation model.

Isaac combines multimodal video understanding, embodied reasoning and robot control into a single, sparse backbone.

[Netris @netris_io]

Netris is the #27 fastest-growing private software company in the US.

Thank you to @Nasdaq for showcasing that milestone in Times Square.

The ranking reflects a record-breaking year: in the last 12 months, we hit 800% ARR growth and added 35+ live deployments (and growing) at many of the largest AI clouds in the world.

A major driver of that growth is the AI buildout. GPU clouds and AI factories are going live everywhere, and deploying the GPUs is only part of the job. Turning a GPU cluster into a cloud business takes NAAM — Network Automation, Abstraction, and Multi-Tenancy — the category Netris built, and the reason operators keep choosing it.

#Inc5000 #AIInfrastructure #NAAM

[Akshat Shrivastava @AkshatS07]

Our next model takes the foundation we have built for frontier perception and visual reasoning and extends into control. Open source model coming soon.

[Armen Aghajanyan @ArmenAgha]

Perception and robotics both run on real-time budgets, and inference speed is usually bounded by model scale.
For our next series of models, we derived new architectures specifically tuned for the physical world. From these speeds alone, you'd never guess the model size.

[ModelFront @modelfront]

ModelFront is featured in @Gartner_inc's 2026 Hype Cycle for Natural Language and Unstructured Content Technologies!

ModelFront is AI to check and fix AI translations or trigger human intervention.

Large translation buyers use ModelFront to scale translation while keeping human quality.

Gartner first covered ModelFront in 2020 when ModelFront created the quality estimation category.

Read more at https://t.co/XfeBQ4OkX0 https://t.co/C20Kloc6iE

[Sriram Krishnan @sriramk]

It is clear open source models and harnesses are having a moment. There's a few factors at work

1/ It is now obvious that you can catch up to near-SOTA performance and do so with a clear training lineage. See:@thinkymachines Inkling launch today.

2/ There are several well-funded, talented teams building open weight models now in the US and abroad. Along with the explosing of other near SOTA models (Grok/Cursor, Muse Spark), it is clear we are going to have a diverse ecosystem of models atleast on coding and agentic use.

3/ Organizations are increasingly looking for control over how their data is used and are willing to trade off some access to frontier level tokens for this control. Organizations and countries are increasingly nervous about the frontier labs potentially competing with them down the road and don't want their data to enable a future competitor.

4/ Open source is a slider: you could bring your own open harness, your evals, your business context and are free to pick and choose your model of choice.

5/ Companies have now actively shifted from "how do we get our people to use tokens" to being uncomfortable with their token cost ballooning without a clear line to revenue.

6/ Geo-politically, countries will be weighing open weight models as a way to get frontier-level tokens inside controlled environments that may not be otherwise possible.

All of this leads to more choice for all of us !

[Ashot Arzumanyan @AshotVC]

@ArmenAgha @AkshatS07 Smart move! Physical is starving for data.

[Ashot Arzumanyan @AshotVC]

@DBuniatyan @activeloop Please elaborate more :)

[Ashot Arzumanyan @AshotVC]

@ArmenAgha @NarenBriar Smells wrong…

[Armen Aghajanyan @ArmenAgha]

1/ Today we're shipping Perceptron Agentic Detection. Describe what you want in natural language, or show one example crop, and an agent grounds it in the image. No fine-tuning, no fixed class list.

[Davit @DBuniatyan]

Coding agents that actually get better the more your team uses them.

Introducing Hivemind: continual learning for AI coding agents.

Hivemind turns the traces from every agent your team runs (Claude Code, Codex, Cursor, Hermes, OpenClaw, Pi) into reusable skills, then pushes those skills across all of them. All on your cloud storage.

Now with SkillOpt built in, your skills get trained: +19.1 points of accuracy in Claude Code, +24.8 in Codex, best or tied on all 52 setups tested.

Open source, one line install.

[Ashot Arzumanyan @AshotVC]

@KanuGulati @bittlingmayer and team at @modelfront have been doing exactly this in translation for years – AI that verifies AI output at scale, in production with F500 co-s.

[Ashot Arzumanyan @AshotVC]

@perceptroninc Love it!

[Ashot Arzumanyan @AshotVC]

@nntaleb I think this totally underestimates the impact of AI agents (currently digital, soon also physical) on LLMs’ access to data and capabilities to learn.

[Netris @netris_io]

Today at @nvidia GTC Taipei, Netris announced the expansion of Netris NAAM (Network Automation, Abstraction, and Multi-Tenancy) on NVIDIA DSX Air across APAC.

APAC is building AI infrastructure at an unprecedented pace. As leading neoclouds, NVIDIA Cloud Partners, AI factories, and sovereign AI clouds are coming online across Japan, Korea, Taiwan, Australia, India, and Southeast Asia, they are turning to Netris as the GPU networking and multi-tenancy foundation for their AI infrastructure.

Netris NAAM and NVIDIA DSX Air are empowering this build-out by enabling AI operators to design, validate, and test their full network stack before hardware ships, allowing them to go live the moment GPUs power on.
Netris’ APAC customer base is growing fast.

@HonHai_Foxconn-backed https://t.co/13UlpnZhgs, the largest GPU cluster and AI supercomputing center in Taiwan, and @firmus_ai, Australia's largest renewable-powered sovereign AI infrastructure, have both standardized on Netris NAAM as the network automation foundation for their AI clouds.

Netris NAAM is the most widely deployed AI network automation and multi-tenancy platform on the market. In the last 12 months, Netris achieved 800% ARR growth, captured 12% of the global neocloud market, and reached 25+ live deployments at many of the largest AI clouds in the world.

📍If you're attending GTC Taipei at COMPUTEX 2026, visit Netris at Booth #N0008, TaiNEX 1 for a live demo.

🔗 Read the Netris and NVIDIA DSX Air APAC expansion press release: https://t.co/N7tcb5Zwnv