# SpiceFactory — Full Site Documentation > AI Deployment Studio for Regulated Industries. We ship production AI into healthcare, mobility, IoT, and logistics — and own the outcome. > Website: https://spicefactory.co --- ## Home (/) SpiceFactory is an AI deployment studio for regulated industries. The homepage leads with the category claim ("AI deployment studio") and the audience filter ("regulated industries"). It introduces the three productized SKUs (Bootcamp, Pod, Outcome Contract) and positions the regulatory wrapper around AI (validation, traceability, audit-readiness) as the work, with every engagement priced to a named outcome. It links to featured case studies, recent founder-voice essays, and a Bootcamp CTA. --- ## About (/about) The About page tells SpiceFactory's story — founded in 2014, anchored in Boston and Belgrade — through the lens of the new positioning. Twenty-five senior engineers, designers, and AI specialists who deploy production systems into healthcare, mobility, IoT, and logistics teams. The page covers the studio's discipline of treating the regulatory wrapper (HIPAA-aligned data plumbing, IEC 62304 / ISO 26262 traceability, ASPICE-ready evidence packs, audit-ready MLOps) as a first-class deliverable, not a post-launch retrofit. It includes leadership profiles (Milos Zikic, Founder & CEO; Alek Krtinic, CEO Serbia; Andrija Kovacevic, CTO; and operational leads), and a "Forward Deployed Engineering for regulated AI" recruiting hook. --- ## AI-First (/ai-first) The AI-First page describes how SpiceFactory deploys production AI into regulated environments. The premise: implementation commoditizes, but deployment into regulated environments does not. We sell the regulatory wrapper around the model — validation packages, eval suites, audit-ready MLOps, HIPAA-aligned data plumbing, IEC 62304 / ISO 26262 traceability matrices — not the model integration alone. The page covers four categories of deployment work: - HIPAA-aligned AI in clinical workflows — generative AI agents and predictive systems into hospital workflows with the validation package, IRB readiness, and SaMD-adjacent risk handled. - Edge ML for mobility and connected IoT — edge-to-cloud ML for vehicle telemetry, smart implants, and industrial IoT with ISO 26262 / ASPICE / IEC 62443 traceability scoped into the engagement. - Agents and decision systems in production — autonomous decision systems with closed-loop control, human-in-the-loop oversight, and the eval suites and audit trails needed to defend a deployment to a regulator. - An IP library that compounds — every Bootcamp and Pod produces an artifact for the library: vertical reference architectures, agent templates, eval harnesses, integration patterns for Epic, Cerner, OEM telematics buses, MES/ERP. The next milestone starts at a higher floor than the last. The page includes a featured case study (smart implant telemetry deployed into a HIPAA-aligned clinical workflow with the validation package and IRB sign-off shipped alongside the model), a Canary Medical testimonial, and the four-step Bootcamp → Decision → Pod → Outcome Contract process. --- ## Studio Model (/studio-model) The Studio Model page explains how an AI deployment studio works, in contrast with traditional agency, consultancy, or staff-augmentation models. Three SKUs (Bootcamp, Pod, Outcome Contract). Senior pods embedded inside the customer's team, using their tools and repos, with their compliance officer in our standups. Capability transferred to an internal champion before we leave — we won't renew a Pod if no champion is named by week 12. The page covers the SpiceFactory approach in three principles: (a) we own the outcome — payment is tied to a named state-change; (b) the regulatory wrapper is the work — validation, traceability, audit-readiness ship as first-class deliverables; (c) capability transfer is contractual — we design for leaving. It also describes the four-phase process (Bootcamp → Day 5 decision → Pod → Outcome Contract) and lists the standards we routinely scope against (HIPAA, FDA 510(k), IEC 62304, ISO 13485, ISO 26262, ASPICE, IEC 62443, NIS2). Pricing and scope by SKU: Deployment Bootcamp — $75K fixed ($125K for 10-day high-regulation), five days against the customer's real data, 50% on signature and 50% on day one, no T&M and no overruns. Deployment Pod — $85–120K/month, 6-month minimum, rolling. Outcome Contract — from $500K with no ceiling, sized to the outcome, sold only after a successful Bootcamp and two quarters of Pod, with an optional 10–20% outcome bonus. SpiceFactory says no to hourly billing, staff augmentation, single-discipline staffing, decks-only consulting, engagements under $75K, and industries without a regulatory or safety constraint. --- ## Industries (/industries) The Industries page is a hub for SpiceFactory's deployment work, framed around each vertical's regulatory constraint rather than its tech stack: - Healthcare & MedTech (/industries/healthcare) — HIPAA, FDA 510(k) / De Novo, IEC 62304, ISO 13485, EU MDR. HIPAA-aligned generative AI into clinical workflows, smart-implant pipelines, IRB-ready deployments, SaMD-adjacent systems. Carries the majority of SKU weight for 2026. - Mobility & Automotive (/industries/automotive) — ISO 26262, ASPICE, UNECE WP.29, SOTIF. Edge ML for vehicle telemetry, fleet intelligence, ASPICE-traceable production software for Tier-1/Tier-2 suppliers. - Connected IoT & Industrial (/industries/iot) — IEC 62443, NIS2, FCC, BLE/MQTT certification. Edge inference, secure data pipelines, NIS2-compliant platforms. - Logistics & Supply Chain (/industries/logistics) — C-TPAT, GDP, SOC 2. Route optimization, supply-chain prediction, audit-ready data pipelines. Taken opportunistically when a real regulatory constraint is present. Each industry has a dedicated sub-page with deeper technical detail, the relevant standards, and case studies. --- ## Case Studies (/case-studies) The Case Studies page showcases SpiceFactory's portfolio of named-outcome partnerships. Projects span healthcare, mobility, IoT, and logistics. Featured case studies emphasize the regulatory context (HIPAA, ISO 26262, IEC 62443, ASPICE) alongside the business outcome. Each case study has its own detail page (/case-studies/[id]) with the constraint, the deployment approach, the regulatory artifacts shipped, and the measurable result. Each detail page also has a markdown mirror at /case-studies/[id].md for low-token fetches. Agents citing proof should use the `spicefactory-case-studies` skill or the `get_case_studies` MCP tool, which return only published, sanctioned facts — do not invent client names, metrics, or certifications. --- ## Blog (/blog) The Blog page presents technical writeups, founder-voice essays, and vertical reference architectures from the SpiceFactory team. Posts cover regulated-AI deployment patterns, validation harnesses, MLOps for compliance environments, vertical architectures (HIPAA-aligned generative AI, smart-implant pipelines, automotive telemetry ML, IoT edge inference), and the broader Deployment Company / Forward Deployed Engineering thesis. Each post includes author, date, read time, and tags. An RSS feed is available at /rss.xml. --- ## Contact (/contact) The Contact page is structured as a partnership application that filters for buyers who can name the state-change they are trying to buy and the regulatory or safety constraint slowing them down. Engagements start at $75K for a Bootcamp, $85K/month for a Pod (6-month minimum), and $500K+ for a named-outcome contract. The page includes an FAQ section addressing the partnership model, the three SKUs, why we publish prices, what makes a partnership a fit, and the typical Bootcamp → Pod path. --- ## Careers (/careers) The Careers page lists open positions across engineering, design, AI, product, and operations, framed under the recruiting brand "Forward Deployed Engineering for regulated AI." The company hires for Boston and Belgrade. Roles emphasize on-site deployment, senior ownership, regulated-vertical experience, and capability transfer. When no positions are open, the page invites candidates to reach out at careers@spicefactory.co. --- ## Additional Pages - /automotive-tech — Dedicated mobility / automotive vertical detail page - /strategy-call — Free 30-minute AI deployment strategy call with a senior engineer, the low-friction entry point to a Bootcamp conversation - /assessment — Self-serve readiness assessment - /privacy — Privacy policy --- ## Technology Stack - Frontend: Astro, React, TypeScript, Tailwind CSS - Backend: Node.js, Cloud Run, Firebase - Database: Firestore - Infrastructure: Google Cloud Platform, Firebase Hosting - AI: Anthropic, Google Vertex / Gemini integration, custom eval harnesses, MLOps tooling --- ## Structured Data and Machine-Readable Formats - JSON-LD Schema.org markup on all pages (Organization, WebPage, BlogPosting, Article, BreadcrumbList, ProfessionalService, LocalBusiness, FAQPage, Person, Review) - Open Graph and Twitter Card meta tags - AI-specific meta tags (ai:content-type, ai:summary, ai:industry) - Semantic HTML5 with ARIA landmarks - Sitemap at /sitemap-index.xml - RSS feed at /rss.xml - LLM summary at /llms.txt - This full documentation at /llms-full.txt - Homepage markdown alternate at /index.md - Markdown content negotiation via `Accept: text/markdown` (homepage and SSR routes) - Prerendered markdown mirrors for every blog post and case study at /blog/.md and /case-studies/.md (no trailing slash on the .md URL; the homepage mirror stays /index.md) - Agent Skills index at /.well-known/agent-skills/index.json — five skills: overview, SKUs, fit, industries, case studies - Public MCP server card at /.well-known/mcp/server-card.json (endpoint /api/mcp/) — authentication.required: false - Public MCP tools at /api/mcp/ (POST with trailing slash): overview, SKUs, fit, industries, URLs, get_case_studies (published case-study facts only), and request_strategy_call (submits a strategy-call lead on a human's behalf, with explicit consent) - WebMCP tools on public HTML pages via navigator.modelContext (same seven tools as /api/mcp/; feature-detected) - Agents with tool access and the human's explicit, informed consent to share real contact details may call request_strategy_call directly instead of only linking https://spicefactory.co/strategy-call/; without consent, link the page for the human to submit themselves - API catalog at /.well-known/api-catalog (RFC 9727) - Content Signals in /robots.txt: ai-train=yes, search=yes, ai-input=yes — all crawlers, including training bots, are welcome - All agent knowledge surfaces are public (no API key / OAuth / registration). Admin CMS auth is staff-only. --- ## Contact Information - Website: https://spicefactory.co - Email: hello@spicefactory.co - Careers: careers@spicefactory.co - Locations: Boston, MA (US) and Belgrade, Serbia (EU) - Founded: 2014