Rime, a Riyadh-based startup founded in 2024, has raised more than $2 million in a seed round led by SEEDRA Ventures, with participation from Athla Investment, Unity Invest Partners and a group of angel investors. The company sells an end-to-end platform that lets enterprises deploy autonomous AI agents inside physical locations — stores, branches and service centres — where decisions are made on-site, on edge hardware, at the moment an event occurs rather than in a remote data centre.
The round matters less for its size than for what it is buying: Rime says it already connects more than 6,000 cameras to AI agents across over 2,000 operational sites, serving retail and service chains in the Gulf. That is not a pilot. It is one of the larger production deployments of agentic AI in the region, and it runs on infrastructure most of those chains already owned.
Key Highlights
- Seed round of more than $2 million, led by SEEDRA Ventures with Athla Investment, Unity Invest Partners and angels.
- Founded in 2024 by Mohammed Almarshadi and Arif Alotaibi; headquartered in Riyadh.
- More than 6,000 cameras connected to AI agents across over 2,000 operational sites in the GCC.
- Processing and decision-making happen on-site via edge computing and AIoT, not in a remote cloud region.
- Funds go toward deepening the Saudi footprint and accelerating platform development, positioned against the National Strategy for Data and AI and Vision 2030.
Details
Rime's pitch is a category its founders call Physical AI: software agents that perceive and act inside an operational environment instead of waiting for a human to open a dashboard. "We are moving into a new era of Physical AI, where systems move beyond screens to interact and make decisions directly within operational environments," said co-founder Mohammed Al Marshadi.
The architectural choice is the interesting part. By running inference on edge hardware at the site, Rime avoids shipping continuous video to a cloud region — which cuts latency to something usable for real-time intervention, cuts bandwidth cost, and sidesteps a meaningful chunk of the cross-border data question that governs camera footage of identifiable people under Saudi Arabia's Personal Data Protection Law. Video that never leaves the premises is a much shorter compliance conversation than video that does.
SEEDRA's thesis is explicitly about existing assets rather than new ones. "Transforming these legacy assets into an active operational intelligence layer represents one of the largest market opportunities today," said Abdullah Al Munif, co-founder of SEEDRA Ventures. Every large retail chain in the Gulf has already spent years installing cameras, point-of-sale terminals and sensors. Almost all of that hardware is used for recording and reconciliation after the fact. The bet is that the upgrade path is software on top, not a hardware replacement cycle.
Impact
For Gulf enterprises, the useful signal here is about sequencing. The dominant regional AI story of 2026 has been capacity — data centres, chips, sovereign compute, HUMAIN's build-out. Rime is a reminder that the returns show up one layer above that, in the unglamorous work of wiring agents into systems a business already runs.
That is also where most deployments stall. A camera network, a point-of-sale system and a staffing roster typically live in three vendors' silos with no shared identifier between them. An agent that can see a queue forming but cannot reach the rota, or can flag a stock gap but cannot open a replenishment ticket, produces alerts rather than outcomes. The integration layer — and the reporting layer above it — is the part that decides whether an AI deployment changes a number on a P&L.
The edge-first design also has a cost profile worth noting for anyone budgeting a similar system. Continuous video analysis in a cloud region is one of the more expensive AI workloads that exists, and it scales with sites, not with usefulness. Moving inference to the site converts a recurring bandwidth-and-inference bill into a mostly fixed hardware cost — a very different case to put in front of a CFO.
Background
Rime's round landed in a week when MENA startups raised roughly $259 million across three disclosed deals, dominated by Moove's $250 million Series C in the UAE. Three of the four deals that week were AI or enterprise-software companies, including HUMAIN's strategic investment into Saudi financial-crime AI firm Mozn — the PIF-backed group's first investment into a Saudi startup.
Saudi Arabia designated 2026 the Year of Artificial Intelligence, and SDAIA reports that Saudi AI companies raised $9.1 billion across 70 deals in 2025. Against those numbers a $2 million seed is a rounding error. But the Kingdom's AI funding is heavily concentrated in infrastructure and a handful of national champions; deals like this one are the thin part of the stack, where operational software actually meets an existing business.
What's Next
Rime says the capital will go toward expanding in Saudi Arabia and accelerating the platform. LEAP opens in Riyadh on 31 August, and the applied-AI end of the Saudi market — agents deployed against real operations rather than model announcements — is likely to be well represented there.
The open question for the category is measurement. Physical AI vendors across the region are selling on deployment counts: cameras connected, sites live. The chains buying it will eventually ask a harder question, which is what changed in shrinkage, conversion, queue times or labour cost. Whoever can answer that with an audit trail owns the renewal.
Related reading: the Gulf's lead in production agentic AI, on-device and edge AI models in the enterprise, and LEAP 2026 opening in Riyadh on 31 August.
If you are weighing an agent deployment against systems you already run — cameras, point-of-sale, ERP, rostering — the integration and reporting layer is usually the constraint, not the model. Talk to us and we will map what your existing systems can actually feed an agent before anyone signs a hardware order.
Source: Wamda