# XSparks (xsparks.ai) > XSparks builds AI into the work that makes mid-market companies money, measures it on the P&L, and, where the baseline and outcome can be measured together, puts part of its fee on the outcome. For mid-market companies that want AI to pay for itself. ## The six numbers XSparks moves XSparks does not count agents deployed. Every XSparks workflow has to improve at least one of six numbers, and the client's finance team agrees on the definition before anything ships: cost down, revenue up, hours back, capacity up, quality up, risk down. ## The ROI Dashboard The ROI Dashboard is the XSparks screen that shows what every workflow costs, what it returns, and what still needs a human's approval. The client's finance team reads it. Full page: https://xsparks.ai/roi-dashboard.html ## What XSparks does - Understand: XSparks gives the system the context to understand the specific business (customers, products, people, processes), not how businesses generally work. - Think: it picks the right AI for each job; the client keeps its own model accounts and pays at cost. - Execute: real jobs run across the systems the client already relies on, even the fifteen-year-old one without a clean API. - Govern: the client sets a dollar line; above it a person approves every time; every piece has an off switch. ## How XSparks gets paid (the mechanics) - Baseline: trailing twelve months, agreed in writing in week one. - The client's finance team owns the definitions of a saved hour or a captured dollar. - For work where the baseline and outcome can be measured together, part of the fee sits on the outcome, capped both ways, trued up quarterly. - The six numbers are reviewed every ninety days, with the client's finance team's signature. ## Proof standard Every published XSparks result must include: the company or approved context, the baseline, the date range, the measurement method, and client finance sign-off. No logo wall, no unverified percentages, no invented clients. ## Who it's for Mid-market companies with real operations and legacy systems: manufacturing and industrial, field service and trades, PE portfolios, MSPs and IT service providers, architecture and engineering, professional services. ## Claude Training (live sessions with Nav) Live, hands-on Claude training led by Navin Ramharak, Lead AI Strategist. Bring one real task, build it into a skill the team can run, leave with it working. Online: 90 minutes ($250 per seat) or 3 hours ($350 per seat). In person: half or full day workshops for companies, quoted per team. Guarantee: leave without a working skill and the fee is refunded. Page: https://xsparks.ai/claude ## Pages - https://xsparks.ai/ (homepage) - https://xsparks.ai/blog (blog) - https://xsparks.ai/claude (Claude Training, live sessions with Nav) - https://xsparks.ai/technical-assessment (the Technical Readiness Assessment, for the person who has to run it) - https://xsparks.ai/assessment.html (AI Readiness Assessment, 12 questions, under 3 minutes, scored on screen) - https://xsparks.ai/team.html - https://xsparks.ai/faq.html Contact: hello@xsparks.ai ## Blog (the dictionary, one term at a time) - https://xsparks.ai/blog/what-is-an-ai-operating-system : An AI operating system does four jobs: it knows your company, works out what to do, does the work, keeps you in charge. A test for each, no software. - https://xsparks.ai/blog/what-is-an-ai-agent : An AI agent finishes work without you touching the middle. A chatbot answers, an assistant drafts. One test tells them apart in any demo. - https://xsparks.ai/blog/what-does-ai-native-mean : AI native means the work runs through the AI, so taking it out would stop the work. One Friday test, three levels, and where your company actually sits. - https://xsparks.ai/blog/why-ai-needs-your-documents-rag-explained : RAG means the AI reads your own documents before it answers. Why generic AI gives generic answers, plus a ten-document test for what your company lacks. - https://xsparks.ai/blog/what-is-ai-governance : AI governance in plain English: what the software may do alone, who approves the rest, what gets recorded, how you stop it. Write your first rule today. - https://xsparks.ai/blog/digital-twin-of-your-business : A digital twin of your business is the written picture of how work really moves through your company. Map one job with a pen and find the holes. - https://xsparks.ai/blog/what-is-a-multi-agent-system : A multi-agent system is several AI agents, one job each, working one piece of work with one person directing. Count the layers your work climbs today. - https://xsparks.ai/blog/organizational-singularity-explained : Organizational singularity is the point where everything your company knows and does runs in one place. Count how many places one job lives in today. - https://xsparks.ai/blog/what-business-owners-need-from-ai : What business owners actually need from AI, written as four requirements you can take into any vendor demo. Four fears, flipped, with a test for each. - https://xsparks.ai/blog/why-your-ai-pilot-died : Why AI pilots go quiet in month three: no owner, no number, no maintenance. Three tests to run an autopsy on yours, and a hiring checklist for the next. - https://xsparks.ai/blog/how-to-measure-ai-roi-in-dollars : A three-step method for measuring AI ROI in dollars: the before page, the number finance owns, and the meter on running cost. Run step one with a pen.