Blog · Solution
AI Agents and Assistants
Writing from Sawhorse Engineering on ai agents and assistants.
- Articles
- 6
- Category
- Solution
- 23 September 2026How to tell whether a task is an AI problemFour questions that separate work worth evaluating with a model from work that belongs in ordinary code.
- 21 September 2026Retrieval over your own documents: what decides qualityMost retrieval systems fail on chunking, metadata and permissions rather than on the model. What to get right before choosing anything else.
- 16 September 2026Guardrails for AI agents that take real actionsAn agent with tools can change your data. The controls that matter are enforced in code, not requested in a prompt.
- 12 September 2026Private and on device models: when they make senseRunning a model inside your own environment removes a data transfer and adds real cost. What actually forces the decision, and what it changes.
- 27 August 2026Evaluating an AI feature before it shipsWithout a fixed set of real cases, the only available verdict on a prompt change is that it seems better. Here is how to build one that is worth trusting.
- 25 August 2026Where AI should hand over to a personThe handover is a design decision with commercial consequences. Getting it wrong in either direction is expensive.
Other solution categories
- Custom CRM
- DMS Integrations
- Document Management
- Learning Management Systems
- ERP Integration and Customization
- POS Integration
- Field Service Management
- Booking and Scheduling
- Customer and Client Portals
- Inventory and Order Management
- Ecommerce Platforms
- CMS and Websites
- BI Dashboards and Reporting
- Workflow Automation Platforms
- Membership and Recurring Billing
- Marketplace Platforms
- Multi Tenant and White Label SaaS
- Mobile Apps and PWAs
- Payment Integration
- Computer Vision and Video Analytics
- Browser Games and Interactive
Let’s talk about the system you need.
Tell us where the current process breaks down. We can map the workflow, examine the integration requirements and define a practical scope for the build.
- A mutual NDA is signed before project details are shared.
- Scope is fixed in writing before any build starts.
- You own the source code, the IP and the documentation at delivery.
- Support plans are available and tailored on a discovery or technical call.