The AI Experimentby Xelec

AI venture of Xelec · United States & European Union

Enterprise AI that survives the operations review.

We are three things at once: a product studio shipping five AI products, an engineering team that takes one use case to a production gate, and a research lab. Xelec, our parent company, builds the hardware — so when inference has to run on the factory floor, we design the board and the model together.

11 days
Fastest pilot to production
94%
Field-level extraction accuracy
38 ms
Median on-device inference
EU + US
Data residency regions

Operating model

One team, three modes of engagement.

Most vendors are a consultancy or a product. Being both is why our pilots reach production — the studio ships the components the engineering work would otherwise rebuild each time.

01

Product Studio

Five shipping products — assistant, document intelligence, revenue CRM, evaluation, edge runtime. Deploy in weeks, not quarters.

  • Deployed in your cloud or ours
  • Per-seat or per-document pricing
  • White-label options

02

Enterprise Engineering

Embedded senior teams that take an AI use case from data foundation to a monitored, audited production system with an owner.

  • Fixed-scope paid pilots
  • Production hand-off with runbooks
  • US and EU delivery hours

03

Research Lab

An internal lab on retrieval, document structure, agent reliability and small-model distillation for constrained hardware.

  • Findings feed the products
  • Published experiment notes
  • Hardware co-design with Xelec

Field notes

What actually stops enterprise AI in Chicago and Munich.

These are the six problems we hear on nearly every first call. Each one has a specific engineering answer, not a workshop.

P-01

The pilot graveyard

Twelve proofs of concept, none in production. No owner, no evaluation harness, no path from notebook to a system operations will accept.

Our answerPaid pilots with a written production gate — pass and we ship it, fail and you keep the findings.

P-02

The data is trapped in documents

Invoices, claims, bills of lading, PDFs, scans, faxes and thirty years of legacy tables. The AI is fine; the input never arrives.

Our answerDocument intelligence and a data foundation layer built before any model work starts.

P-03

Compliance is the real blocker

EU AI Act classification, GDPR transfers, SOC 2 evidence, procurement's data-residency questionnaire. Legal stops the project, not engineering.

Our answerEU and US residency options, model-choice control, DPIA-ready documentation and full audit trails from day one.

P-04

Nobody can prove it is right

A confident wrong answer in a customer-facing or regulated workflow costs more than the whole programme saved.

Our answerEvaluation sets, citation-grounded answers, guardrails and drift monitoring shipped with the system.

P-05

Vendor sprawl and unpredictable spend

Nine AI tools, four contracts, no consolidated view of cost per outcome, and a bill that moves with token prices.

Our answerModel-agnostic architecture, cost telemetry per workflow, and the option to move to smaller in-house models.

P-06

Cloud-only AI cannot reach the floor

Factories, clinics, retail back rooms and field devices have bandwidth, latency and sovereignty limits the cloud ignores.

Our answerOn-device inference on Xelec hardware — the capability most AI vendors simply do not have.

Product studio

Five products in production use.

Licensed, deployable in your region, and reused inside our engineering work — which is how they got hardened.

All products

Procurement answers, up front

Data residency
EU and US regions
Deployment
Your cloud, our cloud, or on-prem
Regulation
EU AI Act and GDPR aligned
Controls
SOC 2 aligned practices
Model policy
Model-agnostic, no lock-in
IP
You own the work product

Parent company

Xelec builds the hardware. That is the part competitors cannot copy.

Twenty years of embedded design, sensor integration and manufacturing sits behind this AI venture. When a workload cannot reach the cloud, we do not apologise for it — we design the device.

  • IoT and embedded prototyping

    Concept to a working, instrumented unit — with an honest read on power, thermal and unit cost.

  • Firmware and electronics product development

    RTOS and bare-metal firmware, schematics, PCB layout, DFM and certification preparation.

  • Model and device co-design

    Distillation targeted at a board we control, so accuracy and envelope are decided together.

  • Manufacturing and fleet operation

    Volume transfer, OTA updates, remote diagnostics and store-and-forward for offline sites.

Macro view of a Xelec-designed embedded board with an AI accelerator module seated in its socket

Public roadmap

What ships next, in the open.

We publish the roadmap because buyers in regulated industries plan in quarters, not sprints.

Full roadmap

Now

Shipping

  • Xelec Eval

    Regression gates and drift monitoring in open beta.

  • Extract — EU form pack

    German, French, Dutch and Nordic invoice and claims layouts.

  • Assist — audit export

    Quarterly DPO-ready query and access reporting.

Next

Following two quarters

  • Xelec Edge general availability

    Two reference boards, fleet OTA, offline-first sync.

  • Revenue — EMEA territory engine

    Multi-currency quota, in-region data separation.

  • Private model hosting

    Customer-dedicated inference in EU and US regions.

Exploring

Research-led

  • Xelec sensor + AI reference kits

    Vision, acoustic and vibration capture co-designed with the parent hardware team.

  • Distilled in-house models

    Client-owned small models trained from production traces.

  • Agent supervision framework

    Reliability patterns from R-04 productised.

Next step

Start with a conversation, not a contract.

Thirty minutes with an engineer who has shipped this before. If there is no fit, we will say so on the call.