Chirag Patel

Product Designer

Writes about AI product UX, agentic interfaces, and keeping a front-end build true to its design.

AI Products & Developer Tools
AI Product UX
B2B SaaS & Workflow Software
Design Engineering

23 pieces by Chirag

Article29 Sept 2026
Accepting AI-generated front-end code: what review is for now
AI generates front-end components in minutes. Accepting one still means checking semantics, behaviour off the happy path, server-side access and rollback.
Read more

Article
3 Nov 2026
Article · 3 Nov 2026
AI design-to-code tools: what they produce, and what a design engineer still adds
Can AI turn Figma designs into production code? Three kinds of tool, what each leaves missing, and when output is a prototype, a start or shippable.
Article
3 Nov 2026
Article · 3 Nov 2026
Credits, limits and overage: billing and usage UX for AI features
Show cost before an action, a meter during use, warnings before the limit, a clear choice at it, and invoice lines a user can explain. Plus admin caps.
Article
3 Nov 2026
Article · 3 Nov 2026
Search UX in Data-Heavy SaaS: Filters, Saved Views and the Zero-Results State
How search and filtering should work in a data-heavy SaaS product: three user jobs, filter rules, saved views and a zero-results state that offers a next step.
Article
27 Oct 2026
Article · 27 Oct 2026
Agent-friendly UI: what we saw when an AI agent operated a SaaS admin screen
What an AI agent did on a SaaS admin screen in our small hands-on test, the design rules we drew from it, and an agent-readiness check for your product.
Article
27 Oct 2026
Article · 27 Oct 2026
AI agent governance: the admin console checklist your enterprise customers will use
Eight questions enterprise IT will ask of your AI agent, from inventory to the switch-off, and what your admin console must show for each.
Article
19 Oct 2026
Article · 19 Oct 2026
Design system tools for SaaS teams, grouped by the job they do
Design system tools for SaaS teams grouped by job, from tokens to adoption analytics. Decide where the source of truth lives before you buy.
Article
19 Oct 2026
Article · 19 Oct 2026
How should an AI product let users choose a model?
Let users compare models on their own input, route automatically when they would rather not choose, and always show which model answered.
Article
19 Oct 2026
Article · 19 Oct 2026
How to add AI to an existing SaaS product: where it helps and what to fix first
To add AI to an existing SaaS product, pick one frequent job whose output people can check and undo, then fix its data, permissions and review first.
Article
18 Oct 2026
Article · 18 Oct 2026
Observability or explainability: which earns trust in an AI product?
An explanation helps one user judge one output; observability shows a team how an AI feature behaves over time. What each should show.
Article
18 Oct 2026
Article · 18 Oct 2026
AI search in a SaaS product: should search and the AI assistant be one input?
Use one input when search and the assistant share a starting point and the mode and source stay visible. Keep two when a wrong answer costs more.
Article
18 Oct 2026
Article · 18 Oct 2026
How to design UX for AI products: a readiness checklist
To design UX for an AI product, first settle the user decision, model role, sources, failure, human control and evaluation. A checklist with next steps.
Article
18 Oct 2026
Article · 18 Oct 2026
Should designers write prompts with ML engineers? Yes, where the output is the product
Yes, where output quality is the product. The designer defines good and bad outputs with examples; the ML engineer owns the model and evaluation.
Article
18 Oct 2026
Article · 18 Oct 2026
When to prototype in code instead of Figma
Prototype in Figma by default. Move to code when the answer depends on real behaviour, data, viewports, AI output, feasibility or speed.
Article
14 Oct 2026
Article · 14 Oct 2026
AI chatbot UX design: best practices for B2B products
A B2B chatbot should answer from the customer's data within each user's permissions, cite sources, admit gaps and hand over to a person with context.
Article
14 Oct 2026
Article · 14 Oct 2026
AI interface design: what changes when part of the interface is generated
When a model generates part of the interface, designers specify components, rules and checks, not screens. What that changes for consistency and testing.
Article
14 Oct 2026
Article · 14 Oct 2026
When Figma and Storybook drift apart, which tools actually help?
No tool decides whether Figma or Storybook is right. See which tools show drift when someone makes a change, what each checks, and when setup pays off.
Article
13 Oct 2026
Article · 13 Oct 2026
Lovable app security: what to fix first before real users arrive
A Lovable app's browser talks to your database, so fix access rules first, then secrets, sign-in and payments, then backups and credit limits.
Article
6 Oct 2026
Article · 6 Oct 2026
Nine AI products that place the human check at the point of consequence
Nine AI products that put checkable sources, preview, approval and hand-off where a wrong output would cost something, and what teams can take from each.
Article
6 Oct 2026
Article · 6 Oct 2026
What are good human-in-the-loop UX patterns?
Eight human-in-the-loop UX patterns for AI products, from evidence-first approval to undo windows and override records, with the research behind each.
Guide
5 Oct 2026
Guide · 5 Oct 2026
How do I keep my React build true to the Figma design?
How to make a React build match its Figma design: define a match, share tokens, map components, build every state, and review what remains.
Article
29 Sept 2026
Article · 29 Sept 2026
Agentic UX: design principles for AI agents that act on a user's behalf
Agentic UX design principles for AI that acts for users: authority per action, approval of the exact change, limits outside the model, half-finished runs.
Guide
15 Sept 2023
Guide · 15 Sept 2023
Marketing automation software design best practices
Make marketing automation software understandable by modelling audiences, signals, rules, actions, authority and campaign state.

Chirag Patel

23 pieces by Chirag

Accepting AI-generated front-end code: what review is for now

AI design-to-code tools: what they produce, and what a design engineer still adds

Credits, limits and overage: billing and usage UX for AI features

Search UX in Data-Heavy SaaS: Filters, Saved Views and the Zero-Results State

Agent-friendly UI: what we saw when an AI agent operated a SaaS admin screen

AI agent governance: the admin console checklist your enterprise customers will use

Design system tools for SaaS teams, grouped by the job they do

How should an AI product let users choose a model?

How to add AI to an existing SaaS product: where it helps and what to fix first

Observability or explainability: which earns trust in an AI product?

AI search in a SaaS product: should search and the AI assistant be one input?

How to design UX for AI products: a readiness checklist

Should designers write prompts with ML engineers? Yes, where the output is the product

When to prototype in code instead of Figma

AI chatbot UX design: best practices for B2B products

AI interface design: what changes when part of the interface is generated

When Figma and Storybook drift apart, which tools actually help?

Lovable app security: what to fix first before real users arrive

Nine AI products that place the human check at the point of consequence

What are good human-in-the-loop UX patterns?

How do I keep my React build true to the Figma design?

Agentic UX: design principles for AI agents that act on a user's behalf

Marketing automation software design best practices