This Frequently Asked Questions (FAQ) page provides answers to common questions about ERP AI Agent services, implementation, pricing, technical requirements, and policies. Please read carefully before engaging our services.
These are the questions we hear most often from executives and operations leaders before they book a first call. If your question is not here, the Contact page is the right next step.
Contact us through our website to schedule a discovery conversation. We start by understanding your specific exception process (AR collections, back orders, quality issues, etc.), current pain points, staff time consumed, and business impact. This initial conversation is no-obligation and helps determine if AI agents are a good fit for your situation.
We build custom AI agents for ERP exception processes including:
Each agent is custom-built for your specific workflow, not generic templates.
We specialize in Acumatica Cloud ERP but work with companies running:
For non-Acumatica systems, we work with your existing implementation partners or IT team for ERP-specific integration requirements.
Standard implementation timeline: 6-8 weeks from kickoff to pilot deployment
Breakdown:
Complex implementations involving multiple systems, extensive customization, or unusual workflows may take longer. We provide specific timeline estimates during project scoping.
We provide training for:
Operations Staff (Agent Users):
Administrators:
Training is delivered through live sessions, recorded videos, and written documentation. Format depends on your preference and team size.
Resource-Based Pricing:
We charge based on dedicated or semi-dedicated resource allocation, not fixed project pricing.
Factors Affecting Cost:
Resource Allocation:
Project Complexity:
Engagement Duration:
What Monthly Fee Covers:
We provide detailed cost proposal after discovery conversation, including recommended resource allocation, estimated timeline, monthly fee structure, and total estimated project cost.
Voice agents make and receive phone calls using natural language processing. They:
Voice agents sound natural and professional, not robotic.
Important Context: We configure and integrate third-party platforms. We do not operate our own infrastructure. Security depends on the platforms used.
Data Access:
Data Transmission:
Platform Security: Third-party platforms we configure have their own security measures:
Our Role:
Your Responsibility:
No Infrastructure Ownership: We do not own, operate, or maintain infrastructure. All infrastructure is provided by third-party platforms or your own systems.
Typical Outcomes:
Important: Results vary by process, current state, and implementation quality. We may indicate realistic estimates during discovery based on your specific situation.
Standard Agreements:
We provide standard agreement templates. You may request modifications, though significant changes may require legal review and could delay project start.
AI Services Scope:
For engagements involving Artificial Intelligence ("AI") services, including but not limited to configuration of Custom GPTs, OpenAI Assistants, AI Agents, and related workflows, our role is strictly limited to providing consulting and resource-based professional services.
Important Clarifications:
Our role is configuration, integration, and consulting only.
Yes. Voice agents can scale to handle increased call volume. Data processing agents handle volume increases without issue. However:
Yes. We work with ERP VARs and implementation partners to provide AI agent capabilities to their clients. Partnership models include:
Contact us to discuss partnership opportunities.
Good Fit Indicators:
Less Good Fit:
We provide honest assessment during discovery. If agents are not right fit, we will tell you.
No. We cannot guarantee specific percentage improvements, time savings, or business outcomes because:
What We Do Commit To:
No. It reads the quality system and the rest of the landscape and correlates them. Your systems of record do not change.
No. Many clients contact us knowing they have operational bottlenecks but unsure whether AI agents are the right solution. Our discovery process helps identify specific processes suitable for agent implementation. We will tell you directly if your situation is not a good fit rather than proposing something that will not deliver results.
No. We build agents custom to your specific process, decision rules, and business requirements. While we have reusable frameworks for ERP integration and workflow logic, the agent decision-making, escalation rules, and communication approach are tailored to your operations.
We work with various ERP platforms through standard API integration. The key requirement is that your ERP system provides API access for:
If your ERP has API capabilities, we can likely build agents for it.
Phase 1: Discovery (Week 1-2)
Phase 2: Development (Week 3-5)
Phase 3: Testing (Week 6)
Phase 4: Pilot (Week 7-8)
Phase 5: Expansion (Month 3-6)
Typical training duration:
We also provide ongoing support for questions that arise after initial training.
Monthly resource fees cover:
Yes. Agents identify appropriately during calls. For example: "This is calling from [Company Name] regarding your order." If asked directly, agents acknowledge being an automated system.
We have found most customers and vendors focus on getting their issue resolved, not on who is providing the information. Professional handling and clear communication matter more than whether it is AI or human.
Important: We do not operate our own infrastructure. Data storage depends on the third-party platforms used in your solution.
Typical Data Storage Locations:
Data Residency: Data residency depends on platform provider policies:
Your Control:
No Long-Term Storage By Us: We do not maintain long-term copies of your operational data. Data exists only in the platforms you license and control.
Primary Metrics:
Secondary Metrics:
We help you define success metrics during discovery and track them throughout implementation and operation.
Services Model - No Product Warranty:
We are a professional services consulting company, not a product vendor. We do not provide product warranties on deliverables.
Ongoing Service Commitment:
As long as you retain and pay for our resource allocation:
Upon Service Termination:
If you discontinue our services (stop paying for resource allocation):
What This Means:
Think of this as ongoing consulting support rather than a warranted product. We remain committed to making the solution work while we are engaged. Once engagement ends, so does our support obligation.
Quality Commitment:
While engaged, we commit to:
This quality commitment applies during the active engagement period only.
Third-Party AI Licenses:
Client shall be solely responsible for procuring and maintaining, at its own cost, all third-party licenses, subscriptions, and usage rights required for AI services, including but not limited to:
We Bear No Responsibility For:
You maintain direct relationships with these platform providers and are subject to their terms.
Business rule changes are normal and expected. You can:
We expect and accommodate business evolution. Agent flexibility is a key advantage over rigid custom development.
Absolutely. We prefer working with your existing implementation partner if you have one. They understand your ERP configuration and can assist with:
We collaborate professionally with implementation partners and value their ERP expertise.
Depending on your situation, alternatives might include:
We are direct about whether agents are the best approach or if alternatives make more sense.
Important Notice:
All website content including use case descriptions, demonstration videos, blog articles, and podcasts is provided for general informational and illustrative purposes only. Such content may be derived from:
This content does NOT constitute:
Your Responsibility:
We strongly encourage you to:
We Disclaim Liability:
For any reliance on website content including:
Decision-making should be based on:
The rules that decide are deterministic, versioned and testable. The model retrieves and explains. The audit trail is append-only.
For an effective discovery conversation, prepare:
ERP automation handles standard transactions well but cannot address company-specific exception workflows. Our agents provide:
Generic ERP features must work for thousands of companies. Our agents work for your company specifically.
No. Agents integrate with your existing ERP system through standard APIs. We do not require system changes, version upgrades, or data migration. Integration typically works with current ERP versions and configurations.
No. Implementation happens alongside current operations, not as a replacement. Pilot deployment starts with limited scope (one product line, one customer segment, specific exception type) so business continues normally while we validate the agent approach.
Only after successful pilot validation do we expand to broader coverage. You control the expansion timeline based on comfort and results.
Ongoing Support Includes:
Support Options:
Support terms and pricing are defined in service agreements.
Yes. Ongoing costs include:
Important: Platform subscription and usage costs are YOUR responsibility and billed directly by platform providers.
These platform costs are outlined during discovery so you can budget appropriately.
Agents recognize escalation triggers:
When these triggers occur, agents transfer immediately to human staff with complete context (who, what was discussed, issue identified). No customers are forced to interact with AI if they prefer human contact.
Important: We cannot make blanket compliance claims because compliance depends on how you implement and use agents.
What We Provide:
Your Responsibility:
If you operate in regulated industry (healthcare, financial services, etc.), discuss compliance requirements during discovery so we can address them in implementation planning.
Typical ROI Timeline:
ROI calculation includes:
Liability Cap: Our total liability for any claims is limited to fees paid by you in the 3 months immediately preceding the claim.
Excluded Damages: We are not liable for:
Your Responsibility:
International Compliance:
If Client uses Services or deliverables outside the United States or Canada, Client shall ensure compliance with all applicable laws including:
Client Responsibility: Client shall indemnify us for any violations arising from such international use. We make no representations about compliance of AI services with laws outside the United States and Canada.
When agents encounter situations they cannot handle:
Agents fail gracefully, not catastrophically. Escalation is feature, not bug.
We recommend waiting until ERP implementation stabilizes before adding AI agents. Reasons:
Once ERP is live and stable (typically 3-6 months post-go-live), we can begin agent implementation.
We do not offer free proof of concepts because:
What We Offer Instead:
The paid pilot serves as proof of concept with real production use.
Agents are software and can make mistakes due to:
When Mistakes Happen:
Your Responsibility:
Our Responsibility:
We do not provide generic product demos because every implementation is custom. Instead:
Your reviewer, exactly as today. The system shortens the queue and shows its reasoning for every finding.
Initial discovery conversation typically takes 60-90 minutes. If the fit looks good, we conduct deeper process mapping (2-4 hours over 1-2 weeks) involving staff who actually handle the exceptions. We document decision rules, escalation criteria, and coordination points before providing a detailed proposal.
Yes. Voice AI agents are our particular focus. Most ERP publishers avoid voice capability because it requires different expertise than traditional software development. We see voice as essential for exception processes involving customer communication (collections, back orders), vendor coordination (quality issues, quotations), and urgent internal alerts.
Voice communication resolves situations faster than email chains and provides better customer/vendor experience.
Requirements vary by agent type but typically include:
ERP and Business Systems:
Third-Party Platform Subscriptions (Client Responsibility):
Our AI agents are built using open source and commercial platform architecture. Client must procure, maintain, and pay for the following platform subscriptions at Client's own cost:
Additional platforms may be required depending on your specific use case.
Open Source Architecture:
Our agent solutions are built using open source architecture and industry-standard platforms. This approach provides:
Client Infrastructure Responsibilities:
Client acknowledges and agrees that:
Cost Planning:
Budget for ongoing platform costs in addition to our implementation and support fees. Typical monthly platform costs range from $100 to $1,000+ depending on:
We provide detailed technical requirements and platform cost estimates during project scoping based on your anticipated usage.
Platform Selection:
While we recommend specific platforms based on proven architecture, Client may request evaluation of alternative platforms. However:
Our approach emphasizes:
Issue Resolution Process:
Issue Categorization:
Resource-Based Engagement Model:
We do not use traditional fixed-price or time-and-materials billing models. These approaches often fail in AI agent projects due to:
Our Approach: Dedicated or Semi-Dedicated Resource
Instead, we provide a dedicated or semi-dedicated engineer/consultant to work exclusively or primarily on your project for a defined engagement period.
How This Works:
Dedicated Resource:
Semi-Dedicated Resource:
Benefits of This Model:
For Client:
For Both Parties:
Typical Engagement Structure:
Phase 1: Initial Implementation (3-4 months)
Phase 2: Refinement & Expansion (2-3 months, optional)
Phase 3: Post-Implementation Support (optional, ongoing)
Implementation Cost Coverage:
Monthly fees cover all consulting and development work. Third-party platform subscriptions (OpenAI, Retell AI, n8n, Supabase) and your staff time are your responsibility.
Investment Range:
Based on typical engagements:
Engagement Terms:
This model works because it aligns our incentives: you want a working solution, we want to deliver it efficiently. Neither party benefits from scope debates or hour tracking.
Yes, completely. You define:
Communication templates are fully customizable and can be updated anytime. The agent speaks with your company's voice, not generic scripts.
We Do Not:
We May:
Your data is yours. We do not monetize it or use it beyond providing services to you.
Our Commitment:
What We Cannot Guarantee:
Realistic Expectations:
Our IP:
Your IP:
Joint IP:
Deliverables Ownership:
For AI services, Deliverables include:
License Grant: Subject to full payment of fees, Client shall have a perpetual, non-exclusive, royalty-free license to use such Deliverables for its internal business purposes.
Our Retained Rights: We retain ownership of all know-how, methodologies, frameworks, and templates used to create such Deliverables, and may reuse them in work for other clients. Your specific business data and proprietary business rules remain confidential.
Standard Reporting:
Custom Reporting:
Reports are accessible through dashboards and can be exported for further analysis.
Ask Us:
Ask Yourself:
Agents Cannot:
Agents Excel At:
Realistic Expectation: Agents handle 60-80% of exceptions automatically, escalating 20-40% requiring human judgment. This is still significant value.
For Prospective Clients:
For Current Clients:
That is the normal starting point. Extractions carry a confidence score, and anything below the threshold goes to a person.
We understand evaluation and budgeting cycles. There is no pressure to proceed immediately. We can provide educational materials, answer questions as they arise, and reconnect when timing is better. Our goal is finding situations where we can deliver measurable improvement, not forcing engagements.
Currently, our voice agents primarily support English. Text-based agent communication (email, portal messages) can support multiple languages depending on your ERP system's language capabilities. If multilingual support is required for your operations, discuss this during discovery so we can evaluate feasibility.
Not necessarily. We handle agent development and integration. Your involvement includes:
If your IT team wants to be involved, that is fine. If you prefer we handle technical aspects with minimal IT involvement, we can do that too.
You need staff participation but not full-time dedication:
We minimize demands on your time while ensuring you stay informed and involved.
Agent modification is normal as your business evolves. Common modifications include:
Minor modifications (threshold changes, template updates) are typically included in support agreements.
Major modifications (new exception types, system integrations, functionality changes) are quoted separately as enhancement projects.
Monthly Resource Engagement:
Standard Terms:
Billing Cycle:
Minimum Commitment:
Payment Methods: We accept checks and ACH transfers.
Engagement Modifications:
Early Termination:
Agents follow the same confidentiality rules as your staff:
You control what information agents can access and share. Sensitive accounts can be excluded from agent handling entirely.
We accommodate reasonable security inquiries within the scope of our consulting practice:
What We Can Provide:
What We Cannot Provide:
We are a consulting practice, not a platform provider. We do not maintain:
Platform Security:
For security audits of the platforms we configure (OpenAI, Retell AI, n8n, Supabase), you should:
Our Security Practices:
We follow industry-standard secure development practices:
Extensive Audit Requirements:
If your organization requires extensive security audits, formal compliance documentation, or on-site security reviews, these may incur additional consulting fees. Discuss your specific requirements during discovery so we can determine feasibility and scope appropriately.
We maintain client confidentiality and do not share detailed case studies without explicit permission. We can provide:
We cannot show proprietary implementations, business rules, or specific client results without permission.
Yes. We provide agreement templates during proposal stage, before you commit. You can:
We want you to be comfortable with terms before starting work.
AI-Generated Outputs:
Client acknowledges that outputs generated by third-party AI platforms may be subject to the platform provider's terms of service regarding ownership and use rights.
No Representations: We make no representation regarding:
Client Responsibility: Client is solely responsible for determining whether AI outputs may be used, commercialized, or claimed as proprietary under applicable law. Consult your legal counsel regarding ownership and use rights.
Yes. Agents operate 24/7 if needed. However:
You control agent operating schedule based on your business needs and customer expectations.
Build Business Case:
Address Concerns:
We Can Help:
Yes. Please review:
These documents are incorporated by reference into service agreements.
Start with a spreadsheet export. No integration is needed to see the workspace running on your own data.
Yes. We strongly recommend pilot approach. Typical pilots focus on:
Pilot duration is typically 90 days. This validates agent effectiveness, customer/vendor acceptance, and ROI before broader deployment.
Our primary focus is companies with $20M+ in revenue in North America. We may consider engagements outside North America on a case-by-case basis depending on time zone compatibility, language requirements, and operational complexity. Contact us to discuss your specific situation.
Agents integrate through standard APIs:
Integration approach is determined during project planning based on your specific systems and requirements.
Yes. Most implementations are conducted remotely through video conferences, screen sharing, and collaborative tools. We have successfully implemented agents for clients across North America without on-site visits.
On-site visits can be arranged if you prefer, but are typically not necessary.
Yes, to an extent. You can:
However, deep technical changes (integration modifications, major logic changes, new functionality) require our involvement. We provide this balance: you control day-to-day operations, we handle technical complexity.
Monthly Resource Engagements:
No Refunds for Current Period:
Work Product Delivery:
Termination Process:
Client Responsibilities Upon Termination:
This policy protects both parties: you get what you paid for through the final billing period, we are compensated for time allocated to your project.
Agents recognize dispute language and escalate immediately to appropriate staff. They do not attempt to resolve:
These situations require human judgment and relationship management. Agents handle routine coordination, not conflict resolution.
Upon Termination:
Your Data in Third-Party Platforms:
Your Data in ERP:
Configuration and Scripts:
Clean Termination: We do not hold data or configurations hostage. Termination is clean and professional. However, you lose the license to use our proprietary methodologies and frameworks.
Dispute Resolution Process:
Governing Law: Agreements governed by laws of jurisdiction where our company is registered.
We prefer resolving issues professionally without legal action, but clear dispute resolution terms protect both parties.
Data Protection for AI Services:
Client Data Ownership: Client acknowledges and agrees that it retains full ownership and control of all data provided or processed in connection with AI services.
Our Limited Role: We shall not be responsible or liable for:
Platform Provider Responsibility: All data protection obligations shall rest exclusively with the Client and the respective third-party AI service providers.
Breach Notification: Client shall be solely responsible for notifying affected parties and regulatory authorities in the event of any data breach, unauthorized access, or security incident involving AI platforms. We shall have no obligation to monitor, detect, or report such incidents, and Client waives any claim against us arising from delayed or inadequate breach response.
Expansion is common and welcomed. Process:
Having one successful agent makes additional agents faster and less expensive to implement.
No. The AI agents we build are not designed to replace employees.
They are designed to take over repetitive, coordination-heavy exception work that consumes time but does not require human judgment, context, or accountability.
In most ERP environments, exceptions are handled by capable people spending large portions of their day on:
AI agents handle those tasks consistently under defined rules.
Employees remain responsible for:
In Practice:
Companies use AI agents to reduce overload, not headcount. Teams typically redeploy time toward:
If your goal is workforce reduction, this is likely not the right solution. If your goal is to help experienced staff spend more time on work that actually requires experience, AI agents are often a good fit.
Yes. This FAQ page, along with all other content on the ERP AI Agent website, constitutes the terms under which services are offered and delivered.
By engaging with us for discovery conversations, requesting proposals, or executing service agreements, you acknowledge that you have read and agreed to be bound by:
These terms apply regardless of which affiliated entity ultimately delivers services and issues invoices. Service agreements will incorporate these website terms by reference.
Terms in effect at the time of initial engagement govern the service relationship.
Yes. Decide this before the build, not after.
ERP AI Agent is a marketing and business development initiative. Actual consulting services, implementation work, and invoicing may be provided by an affiliated professional services entity or another qualified service provider within our network.
Service Delivery:
Why This Structure:
This approach allows us to:
Terms and Policies:
Regardless of which affiliated entity delivers services:
Client Acknowledgment:
By engaging with ERP AI Agent for discovery, proposals, or contracting discussions, you acknowledge and agree that:
For Clarity:
When you receive a formal service agreement or invoice, it may reference:
This is normal and expected. The delivering entity is authorized to perform services under the ERP AI Agent brand and methodologies described on this website.
Questions About Service Entities:
If you have questions about which entity will deliver services, this will be clarified during the proposal and contracting phase, before you commit to any engagement.
Agents integrate through standard API patterns that remain stable across ERP versions. Minor system updates typically do not affect agent operation. Major version changes or system replacements may require integration review and potential updates, but not complete agent rebuilds.
We can monitor for ERP updates and proactively address any integration impacts. This is covered under ongoing support agreements.
Client-Hosted Infrastructure:
The AI agent solution is deployed and hosted in Client's own environment (Client's cloud account, self-hosted servers, or Client's selected infrastructure). We do not host, operate, or control the deployed agent.
Upon Service Termination:
If Client decides to discontinue our consulting and support services:
Client's Continued Use Rights:
Client may continue to use the deployed agent solution after service termination:
Code and Intellectual Property:
Client's Rights: Subject to full payment of all fees, Client receives a perpetual, non-exclusive, royalty-free license to use, operate, and modify the delivered code and configurations for Client's internal business purposes. Client owns the deployed instance and may continue using it indefinitely.
Our Rights: We retain ownership of all methodologies, frameworks, techniques, know-how, and reusable components used to create the solution. We have the unrestricted right to:
What This Means: Client owns the specific deployed code for Client's use. We own the right to create similar solutions for others using the same methodologies and frameworks. Neither party may restrict the other's rights as described above.
Infrastructure Wind-Down:
If Client chooses to decommission the agent:
No Ongoing Obligations:
After service termination:
Professional Transition:
We will assist with transition planning during the notice period to ensure Client understands ongoing operational requirements and can operate the agent independently or with alternative support providers.
We do not offer internal financing. However:
Contact us to discuss options if budget is a constraint.
Voice agent capacity depends on:
Typical voice agent setup handles 40-60 calls daily per agent instance. We can deploy multiple instances for higher volume.
AI Output Disclaimer:
AI systems are probabilistic in nature and may generate inaccurate, incomplete, biased, or otherwise unexpected outputs.
No Warranties: We make no representations or warranties, express or implied, regarding:
Client Validation Required: Client assumes sole responsibility for:
No Guarantees on AI Behavior: We cannot guarantee how AI systems will respond to specific inputs or that outputs will meet your expectations.
Yes, but usually in subtle and positive ways.
AI agents change how work flows, not who is responsible.
Typical Changes:
Roles do not disappear. They become more focused.
Examples:
These changes are usually welcomed by teams once they see how the agent works in practice.
It depends on your sources, volumes and hosting. We discuss it with your numbers in a working session.
Yes. Many of our clients have customized ERP implementations. As long as:
We can build agents that work with your customized environment. We may need additional discovery time to understand your customizations.
Agents support multiple communication channels:
You define channel preference by customer segment. VIP customers might get calls, standard customers get email, etc. The agent adapts to your communication strategy.
No Service Levels:
We do not provide, and expressly disclaim, any service-level commitments, performance guarantees, or uptime warranties with respect to:
Platform Provider SLAs: Service levels, if any, are provided directly by the platform providers under their own terms. We have no control over platform availability or performance.
Best Efforts Only: We provide best-efforts support for issues within our control (configuration, integration logic) but cannot guarantee resolution of platform-level issues.
Client Indemnification:
Without limiting other indemnification obligations, Client shall indemnify, defend, and hold harmless our company from and against any claims, damages, or liabilities arising out of or related to:
AI Output Reliance:
Compliance Violations:
Intellectual Property:
Business Impact:
This indemnification is in addition to liability limitations stated elsewhere and survives termination of services.
Book a 30-minute discussion about your specific situation. No sales process. No commitment required. Just a conversation about whether AI agents make sense for your ERP environment.
