# Noesis: frequently asked questions Reviewed: 2026-09-28 Canonical page: https://noesis.tech/faqs ## What does Noesis do? Noesis builds AI agents and connected business systems to help companies sell more, run better and serve customers better. We embed trained engineers, analysts and consultants with your team to understand how work happens, identify a useful problem and build the solution in that working context. [Related information](https://noesis.tech/about) ## Which businesses are a good fit for Noesis? Noesis suits operating businesses with recurring technology needs, disconnected systems or customer journeys that need improvement. A good engagement has a business owner, access to the relevant systems and a measurable problem. An unclear mandate or a request to automate everything at once is a poor starting point. [Related information](https://noesis.tech/how-we-work) ## What does an embedded AI engineering team do? An embedded team works with the people who use and manage the process. Engineers, analysts and consultants observe the workflow, review exceptions, build integrations and test solutions with users. Noesis combines that delivery work with an agreed roadmap and outcome reviews. Team capacity and working arrangements are set in the engagement. [Related information](https://noesis.tech/how-we-work) ## Does embedded mean the team works in our office? Embedded describes how closely the team works with your organisation; it does not, by itself, promise a full-time on-site team. We agree the on-site, remote or hybrid arrangement, access needs and review schedule before work starts. Noesis is based in Mumbai and works with businesses across borders. [Related information](https://noesis.tech/how-we-work) ## What business outcomes can an AI agent improve? Useful targets include faster enquiry handling, fewer manual handoffs, better product discovery and less time spent finding business information. The right target depends on your starting point. Noesis agrees a baseline and success measure before building; an agent launch alone is not evidence of revenue growth or cost savings. [Related information](https://noesis.tech/outcomes/systems) ## When should we use an AI agent instead of ordinary automation? Use fixed rules for predictable work with clear inputs and outputs. An AI agent can help when work involves language, varied requests or selecting a next step from context. Many useful systems combine both: AI interprets the request, software checks the rules, and a person approves actions that need judgment. [Related information](https://noesis.tech/outcomes/systems) ## How is an AI agent different from a chatbot? A chatbot is a conversational interface. An agent can also use approved tools to retrieve information or carry out a task. Those capabilities depend on the integration, permissions and workflow built around it. Noesis defines the actions an agent may take and where it must ask a person for help. [Related information](https://noesis.tech/outcomes/product) ## Can Noesis build an AI assistant for WhatsApp? Yes. Noesis builds conversational assistants for business uses such as sales support, product questions and routing enquiries to a person. WhatsApp delivery also needs the appropriate business account, platform integration and current messaging permissions. The scope should cover human handoff and ongoing review, as well as the conversation itself. [Related information](https://noesis.tech/work/truediamond) ## Can an agent work with our existing CRM, ERP or ecommerce platform? Where access and integration options permit, we can connect an agent to existing business systems. We first check the available APIs, data quality, access controls and limits. Support for a specific platform is confirmed during discovery. We do not assume that every system offers a safe or reliable integration. [Related information](https://noesis.tech/outcomes/systems) ## Do we need to replace our current software? Usually the first step is to assess what can be improved within the systems you already use. A better integration, a clearer process or a focused interface may be enough. Noesis recommends a replacement only when the business case and constraints support it; a new platform is not the default starting point. [Related information](https://noesis.tech/outcomes/systems) ## How does Brainstormer fit into a Noesis engagement? Brainstormer is our business-agent development platform. It can provide a reusable foundation for agents that work with business knowledge, workflows and communication channels. Its use depends on the problem and agreed architecture. Platform subscriptions, implementation work and ongoing delivery capacity should be made explicit in the commercial scope. [Related information](https://noesis.tech/about) ## How should we choose the first AI use case? Start with one frequent task that has a clear owner, usable data and a result you can measure. Map the current process, including exceptions and manual effort. A good first use case is narrow enough to test with real users and valuable enough to justify maintaining it after the pilot. [Related information](https://noesis.tech/how-we-work) ## Can we start with a pilot before a longer engagement? Yes. We can discuss a defined audit, workflow or product pilot to establish fit. The scope should state the users, inputs, allowed actions and success criteria, along with what happens after the pilot. A useful pilot tests a business assumption and gives both teams evidence for the next investment. [Related information](https://noesis.tech/start) ## How long does it take to build and launch an AI agent? There is no single timeline that applies to every agent. The schedule depends on the workflow, integrations, data access, testing and approval needs. Noesis agrees a delivery plan after reviewing those dependencies. A limited pilot and a production system with monitoring and support are different scopes. [Related information](https://noesis.tech/how-we-work) ## How will we know whether the system is working? Agree the baseline and success measure before development. Depending on the use case, track task completion, response time, manual effort, qualified enquiries or exception rates. Review quality and failures alongside speed. Compare results over an agreed period and keep changes traceable so improvements can be assessed. [Related information](https://noesis.tech/how-we-work) ## Can we require human approval before an agent takes action? Yes. Approval steps can be part of the workflow. We define which actions are automatic, which require approval and which are prohibited. An agent that answers a product question needs different controls from one that changes an order or record. Testing should cover failed actions and escalation to a person. [Related information](https://noesis.tech/outcomes/product) ## What access and data will Noesis need? We first identify the minimum data and system permissions needed for the agreed task. Discovery may require process examples, representative records and help from a system owner. Production access, data handling, hosting and retention must be agreed for the project. Do not send passwords or sensitive customer records through the website enquiry form. [Related information](https://noesis.tech/start) ## Can an AI system run in our own cloud or on our premises? Hosting requirements should be assessed during discovery. The choice depends on the model, integrations, hardware, data restrictions and support needs. Noesis can discuss a suitable architecture, but private or on-premise operation is not included by default. Hosting a system privately also does not automatically make it offline. [Related information](https://noesis.tech/start) ## How much does a Noesis engagement cost? Retained partnerships start from approximately ₹1.5 lakh per month. This is a starting point for agreed capacity, not a fixed price for every AI project. Scope, skills, responsibilities and term are agreed before work starts. Third-party software, media and infrastructure are priced separately. [Related information](https://noesis.tech/how-we-work) ## What does the retainer include, and is the work unlimited? The retainer provides finite, agreed delivery capacity with priorities that can change as the business learns. The scope can combine engineering, analysis, design and specialist input. It does not mean unlimited requests or a dedicated person for every skill. Major separate builds or extra capacity need a separate agreement. [Related information](https://noesis.tech/how-we-work) ## Who maintains the system after launch? An ongoing partnership can cover monitoring, issue resolution and further improvements within the agreed capacity. The scope must name the owners of operations, support, infrastructure and third-party services. A one-time build should not be assumed to include ongoing support; those responsibilities are agreed before handover. [Related information](https://noesis.tech/how-we-work) ## What should we send Noesis to start a conversation? Tell us about your business, one outcome you want to improve, the current process, the systems involved and your timing. Include a budget range if available. Use the enquiry form for an initial discussion. Submitting an enquiry is not a purchase, a confirmed booking or acceptance of commercial terms. [Related information](https://noesis.tech/start)