Xpander, founded by ex-AWS engineers, closed a $7.5M Seed led by Pico Venture Partners (with Emerge Ventures, Samsung Next, and SeedIL) to scale its vendor neutral enterprise AI agent platform and newly launched Omni agent.
Xpander raised a $7.5 million Seed round led by Pico Venture Partners, with participation from Emerge Ventures, Samsung Next, and SeedIL. Proceeds will accelerate market penetration of its vendor neutral, all in one AI enablement platform.
The company was founded in 2024 by former AWS principal engineers David Twizer (CEO), Moriel Pahima (CTO), and Ran Sheinberg (CPO). Their experience guiding large enterprises through complex, multi year cloud migrations directly shapes Xpander’s approach: treating the shift to AI native operations as an infrastructure and governance challenge rather than a pure model or application problem. Headquarters are in San Francisco; the company maintains strong Israeli roots and is SOC 2 Type II certified and GDPR compliant.
Available data indicate earlier capital of roughly $3M pre seed (2024) plus a $3M pre seed extension (2026, involving SeedIL and others), bringing cumulative funding to approximately $13.5M after this Seed. Headcount figures vary across trackers (from low double digits in some databases to claims in the 51–200 range). The company already serves global enterprise customers in retail, manufacturing, financial services, technology, and government, though specific logos and revenue metrics have not been disclosed.
What is Xpander’s technology?
Xpander positions itself as infrastructure for enterprise AI agents rather than another point solution or model locked stack. Core elements include:
- A universal agent harness: a model, framework, and cloud agnostic runtime that executes agents as portable workloads inside the customer’s own environment (AWS, GCP, Azure, VPC, on-prem, or fully air gapped). It supports any model (frontier or fine tuned) and frameworks such as LangChain while preserving governance.
- Strong emphasis on governance and sovereignty: per user permissions, end to end authentication (agents identify as the invoking human via OIDC), full audit logging of every tool call, spend attribution, credential injection from a vault (so models never see secrets), and centralized lifecycle management.
- Multiplayer/collaboration features: shared conversations, published Agentic Apps, and integration into existing tools (Slack, Teams, ChatGPT, Claude, etc.).
- Omni, introduced alongside the funding: an “agentic Forward Deployed Engineer” that builds, runs, and optimizes agents from plain language. It supports complex multi turn, multi tool, long horizon tasks and collaborative multi agent workflows. Omni scored 90.9% on the GAIA benchmark; methodology and results are published on GitHub.
Both the platform and Omni are generally available (cloud version with free credits and self hosted options).
Xpander targets the well documented gap between AI experimentation and production maturity. Citing McKinsey data, the company notes that 88% of organizations use AI in at least one function, yet only ~1% describe deployments as mature and roughly two thirds remain stuck in pilots. Enterprises face a binary choice: lock into a single vendor’s full stack or manage fragmented point solutions that create operational and governance nightmares. Shadow AI (desktop agents used with or without IT approval) is already widespread.
Xpander’s bet is that the winning layer is a neutral, governed control plane that lets organizations own their AI stack while deploying agents safely at scale. This mirrors the cloud migration pattern the founders lived through at AWS: infrastructure and operational maturity, not just capability, determine who captures the value. Investor commentary reinforces the thesis, Pico’s Tal Yatsiv highlighted the scarcity of true enterprise wide platforms amid a flood of point solutions and cited the team’s experience plus early customer traction as decisive.
Recommended: Malachyte Raises $10 Million In Seed Funding Round
The round is modest by late 2020s AI standards yet strategically timed. It funds commercialization and product expansion just as enterprises move past the first wave of generative AI pilots and confront the harder problems of production agents, multi model fleets, compliance, and cost control. Vendor neutrality is deliberately “unfashionable” relative to closed stacks from major labs, but it aligns with the emerging reality of plural model usage and the need for independent governance that no single frontier provider is incentivized to build for rivals.
Strengths include the founders’ enterprise pedigree, early cross industry customers, a clear product differentiation around portability + governance, a strong public benchmark result for Omni, and readiness (platform + agent already live). Risks typical of the category remain: converting pilots into durable production deployments at scale, competing against both hyperscaler offerings and well funded agent platforms, and proving that the governance layer can keep pace with rapidly evolving agent capabilities without introducing friction.
The $7.5M Seed positions Xpander as an infrastructure contender focused on making AI agents operationally viable and controllable inside real enterprises, rather than another application or model play. The combination of AWS honed enterprise instincts, a neutral architecture, and an immediately usable high performing agent (Omni) gives it a coherent wedge into a large and still unsolved market problem.
Please email us your feedback and news tips at hello(at)techcompanynews.com
