The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Before adding AI to your SaaS, write the failure contract.Most AI feature briefs describe what should happen when the model works.Production problems begin when it doesn’t.Before adding AI to an existing SaaS product, I think the team should define: 1. LatencyHow long can the user wait before the product falls back, continues asynchronously, or offers another path?2. UncertaintyWhat happens when the output is incomplete or confidence is too low? Does the product ask for clarification, request human review, or refuse the action?3. PermissionsWhich records can the model access for this user, role, and organization? AI should inherit the product’s access rules rather than bypass them.4. Action boundariesCan the AI recommend, draft, or execute? Sending an email, changing a payment state, and updating a customer record should not share the same approval rules.5. CostWhat is the usage limit per user or tenant? Which model or workflow becomes the fallback when a request is too expensive?6. RecoveryCan the team see which model, prompt, data, and tool call produced the result? Can a failed workflow be retried without duplicating the action?Connecting an API is usually the easy part. The real product work is deciding how the feature behaves when reality does not match the demo.If you are building AI into a product, which part is hardest to define: latency, permissions, human approval, cost, or recovery?
Recovery is the one I find hardest to define. If an email goes out and the next step fails, I want the retry to know it was sent so the customer does not get it twice.
Exactly. The important distinction is between retrying the workflow and replaying its side effects. I’d persist an idempotency key and an outbox or event record before advancing to the next step, so a retry can reuse the existing email result instead of sending it again. The...
Recovery, for me. Latency and cost you can put a number on, but deciding what is safe to retry for each action is a real product decision. Idempotency keys on every AI-triggered action are the first thing I'd add, so a retry never sends the same email or updates the same record twice.
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
Gridora is a renewable energy SaaS landing page created for grid operators and owners of solar and wind assets.
The goal was to make complex energy optimization feel clear, calm, and easy to act on. The page explains how teams can forecast peak loads, compare storage and market scenarios, reduce curtailment, and keep more clean energy on the grid.
The visual system combines large typography, soft greens, open landscape imagery, thin interface lines, and lightweight data modules to create a product story that feels both technical and approachable.
B2B SaaS website design for Proof — an identity platform that lets AI agents act on behalf of real people, built to convert enterprise security buyers.
Proof issues cryptographically verified credentials: a human verifies once, and their AI agent is then authorized to sign, notarize or transact for them. The site had to sell a category most buyers had never heard of — agent authorization — to security, compliance and engineering leads at banks, law firms and hospitals, and move them into a sales conversation instead of a bounce.
The hero says it in five words — "Verify one. Authorize for everything." — next to a live credential card, so an abstract protocol becomes something you can see in two seconds. The page alternates near-black, white and electric blue, and every switch marks a new argument: the risk, the mechanism, the network, the platform, the API. That rhythm keeps a 14-section B2B SaaS website scannable in a single scroll.
I built 20+ Lottie animations for the parts copy can't carry — credential signing, the authorization handshake, the global network map — kept light enough not to cost load time. For imagery I wrote a style recipe for AI generation (one lens, one lighting setup, one grade) so the finance, legal and healthcare scenes read as one commissioned shoot rather than stock AI. Technical buyers get a real code block with real API calls, because that section is what they scroll to find.
14+ sections, desktop and mobile designed separately, a component library, Lottie files ready for the dev team, and a written art-direction guide for generating new imagery in-brand.
Project Title:
Unifai — AI-Powered HR Operations Platform
Short intro:
A smarter HR experience built around conversation, context, and actionable workforce insights.
Project description:
Unifai is an AI-powered HR platform designed to reduce the friction involved in navigating complex employee and compliance data.
Rather than forcing HR teams to move between dashboards, reports, employee records, and filters, the experience introduces an AI assistant as a central interaction layer. Users can ask questions naturally, investigate issues, access employee context, and move from an insight to the relevant information without breaking their workflow.
The product experience combines conversational AI, compliance monitoring, employee insights, attendance data, onboarding, and workforce analytics within a unified interface.
Our design approach focused heavily on information hierarchy. HR platforms can quickly become dense, so the interface uses progressive disclosure, contextual recommendations, focused data visualisation, and an AI-first interaction model to keep complex information approachable.
The result is an enterprise experience that feels less like navigating HR software and more like having an intelligent assistant sitting on top of your workforce data.
Services:
Product Design · UX/UI Design · SaaS Design · AI Product Design · Dashboard Design · Design System
Industry:
HR Tech · Enterprise SaaS · Artificial Intelligence
This is impressive, detailed and on point.