Overview: Deltia.ai provides real-time, video-based process analytics for manual and semi-automated assembly lines. The system converts camera feeds into metadata and KPIs to detect bottlenecks, measure cycle times, map flows, analyze root causes, and monitor deviations from standard operating procedures. It is positioned as an “AI co‑pilot” for process engineers to improve productivity, quality, and layout/flow (examples noted include walking-route optimization and station-level analysis). Key capabilities: Bottleneck detection and alerts; cycle-time and station metrics; process anomaly detection; walking-route/shop-floor optimization; root-cause analysis and best-practice extraction. Delivery options include a cloud SaaS dashboard plus edge deployments with models running on on‑prem/edge hardware. Privacy measures and anonymization are claimed for EU/US deployments. Typical time-to-insight per press/partner coverage is dashboards available within approximately two weeks after sensor install. Technology & deployment: Uses computer vision models trained and deployed on edge devices. Public partner/press material highlights NVIDIA Metropolis and Jetson AGX Orin with TensorRT model compression for efficient edge inference. Deployment options referenced include Cloud SaaS, Edge on customer hardware, Deltia hardware, or Hybrid. Customers & traction: Press and partner coverage (e.g., NVIDIA blog) reports customers such as Viessmann and ABB and quotes performance uplifts (examples up to ~20–30% per line). Reported seed funding of €4.5M (Cavalry Ventures lead; Merantix and several angels/operators participated). Company founded in 2022; founders/co‑founders named in coverage are Maximilian (Max) Fischer and Silviu Homoceanu. Headquarters: Berlin. Company/hiring: Careers page indicates a small, mission-driven startup with hybrid (Berlin + remote) model, invites speculative applications and emphasizes ownership and product-focus AI. Legal/contract summary (from Terms & Conditions): Agreement structure covers Order + Service Agreement (pilot/live) + Features/Support/T&Cs; Order prevails on conflict. Targeted to business customers (legal entities); Deltia can request proof of status. Term is generally one year with automatic renewal; pilot phase billed one‑off, live phase billed as annual base fee plus customization/hardware and possible overage charges. Pricing adjustments may occur annually (examples reference CPI or ~5% caps). Data use: customer data is collected and results provided via Dashboard; Deltia reserves rights to use learnings/training data for product development and AI training but commits not to share raw customer data with other clients. Deltia retains IP over models, concepts and training data. Standard contract provisions referenced: liability limits, indemnities (esp. IP claims), confidentiality, feedback licensing, assignment restrictions. Governing law: Germany; jurisdiction at Deltia’s registered office. Gaps / missing public information: No public pricing plans or free-trial details (the /pricing URL returned 404/placeholder). Several pages including /features and some docs paths returned 404/placeholder. No public SLA, no public security certifications (e.g., SOC2) or detailed DPA found on the site. No public API documentation or deep technical integration guide was located on the public site. Recommended next steps: (1) Request from Deltia official pricing plans and written quote (pilot vs live; hardware and customization costs), SLA and uptime/maintenance terms, DPA and details on anonymization/retention/access controls, security certifications/audit reports (SOC2, ISO27001, penetration tests), integration checklist and hardware requirements (camera specs, network, edge device throughput), and case studies or references (Viessmann/ABB) with before/after metrics. (2) Book a demo to see dashboard, ingestion pipeline, and example KPIs. (3) If procurement/legal review is needed, have counsel review IP/data-use clauses and confirm restrictions around model training and derived insights. (4) If evaluating a pilot, request a scoped pilot Order including acceptance criteria, rollout plan, and KPIs to measure. Sources referenced in the review: NVIDIA blog (Metropolis / Jetson coverage), SaaS press coverage for seed funding, LinkedIn posts and Crunchbase/Tracxn entries for company profile and hires. All above statements reflect only the content found during the site and related source review; no additional claims beyond cited materials were made.