Three SaaS tools you can build with AI instead of renting
In this issue
Skip the LinkedIn detective work: see the buyers showing intent, live on Nov 5
Account-level intent data plays a game of Clue: someone at Acme, somewhere in the building, with a budget. Sales still has to guess who.
Influ2’s new release, Account Discovery That Starts With Who’s Showing Intent, names the suspect first. It:
- Finds named people showing intent across your entire market
- Surfaces the accounts they work at
- Fills out the rest of the buying group
Mark Thursday, Nov 5: the live first look starts at 12 PM ET.
You’ll see each use case step by step, from a huge ICP list with no obvious starting point to accounts showing intent that slipped past you.
Headliner alert: Scott Brinker, the “Godfather of Martech,” joins Indico Data’s Caroline Hsu and Influ2’s Anna Tsymbalist and Kaela Clark.
Can’t make it live? Register anyway, and Influ2 will send you the recording.
Question: How many SaaS subscriptions do you pay for every month?
Another question: How many of their features do you use?
With recent AI improvements, you can stop paying for the whole package and build only what you actually need.
In October, developer Brandon Thomas showed off ArtCraft, seven Adobe-inspired apps he built with Claude’s help. Among them: his own takes on Photoshop and Premiere Pro.
Paid software may survive this one. But for SaaS companies, ArtCraft could be a warning sign.
Using AI to rebuild existing software even has a name: Reverse Engineer Anything (REA). The trend shows just how much AI is changing the game.
And you don’t even need to recreate an entire app to benefit.
So we picked three popular platforms and worked out how to recreate their most useful features with AI.
Build your own form builder
Tools like Typeform make great forms. But do you really need another subscription just to ask your customers a few questions?
With AI, you can build your own form builder, complete with conditional logic, progress bars, design customization, and CSV exports.
The best part? The first prototype needs nothing but HTML, CSS, and JavaScript. No extra libraries or complicated setup.
Do it now:
- Choose the essential features: Start with a question builder, step-by-step forms, and design customization.
- Generate the app with AI: Describe the features you need and ask for the complete HTML, CSS, and JavaScript code.
- Add conditional logic: Make sure the next questions change based on previous answers.
- Set up response tracking: Save responses in the browser’s local storage, then add a simple dashboard and CSV export.
- Test the finished form: Check how it works on mobile, verify the responses, and test every possible question path.
Use our starter prompt:
Build a professional, fully functional web application inspired by the core features of Typeform, using only HTML, CSS, and vanilla JavaScript. Read more…
Check out our example here.
Pro tip: Want a more professional tool with user accounts, a database, and shareable forms? Use React for the interface and Supabase to store the data. Then add authentication, user permissions, and proper data security.
Treat the HTML, CSS, and JavaScript version as a prototype. It isn’t ready to handle real customer data. And yes, you can build all of this with AI, too.
Build your own appointment scheduling tool
“Does Tuesday work?” “What about Thursday?” “How about Friday at 3 PM?”
Scheduling a single meeting can turn into an endless back-and-forth of emails. Calendly and similar tools fix that by letting people book a time from your open slots.
But you can also build its core features yourself.
A calendar, available time slots, booking forms, and automated confirmations? Instead of paying for another subscription, let’s give AI the task and see what it can build.
Step by step:
- Choose the essential features: Start with an interactive calendar, time slot selection, and a customer booking form.
- Generate the app with AI: Describe the features you need and ask for the complete HTML, CSS, and JavaScript code.
- Set up availability: Add working days, meeting hours, appointment durations, and a way to block unavailable time slots.
- Add booking functionality: Let users choose a time slot, enter their details, and receive a confirmation. For the prototype, you can store bookings locally.
- Scale it up if needed: To handle real customers, use Next.js (a React-based framework) and Supabase, connect Google Calendar, and add email notifications. Make sure the database itself blocks double bookings.
Grab our starter prompt and customize it to fit your needs:
Create a beautiful, modern, fully interactive appointment scheduling application that demonstrates how AI can recreate the essential functionality of a paid SaaS scheduling platform. Read more…
Check out our example here.
Build your own A/B testing tool with AI
The new headline looks great. So does the CTA button.
But are customers actually clicking more?
Get AI to build your own A/B testing tool, one that splits traffic, tracks conversions, and displays results on a clear dashboard.
You’ll still need enough data to draw reliable conclusions, though.
Playbook:
- Choose what to test: Start with headlines, CTA buttons, colors, or sections on a landing page.
- Generate the app with AI: Describe the features you need and ask for the complete HTML, CSS, and JavaScript code.
- Set up traffic splitting: Make sure the tool randomly assigns visitors to version A or B and remembers which version each visitor got.
- Build a results dashboard: Add visitor counts, conversion rates, and a chart comparing both variations.
- Test everything: Check that both versions display correctly, visitors keep their assigned version, and the calculations work with sample data.
See how we created our starter prompt:
Users must be able to create experiments, customize landing page variations, assign visitors to different versions, record conversions, and analyze results through a professional dashboard. Read more…
Check out our example here.
Build it, don’t rent it
So… in a few prompts you can build:
- Your own form builder with conditional logic and response exports.
- A scheduling tool where customers book available time slots themselves.
- Your own A/B testing tool and compare the results.
And that’s just the three we picked.
The takeaway? You don’t need to recreate entire SaaS platforms.
Sometimes, a few carefully chosen features, a little help from AI, and one solid prompt are enough to build a tool that fits your needs.
And that monthly subscription? It might not survive your next renewal date.
Become an AI expert in just 5 minutes a day
If you're a decision-maker at your company, you need to stay ahead of AI without spending your morning in webinars, reports, and half-finished podcasts. There is a faster way:
This daily newsletter condenses the latest and greatest AI developments into a 5-minute read. They read the noise, pull out the signal, and explain what actually matters for work, strategy, and the tools your team is already using.
Trusted by 700,000+ readers at Google, Meta, Microsoft, and beyond.
AI DESIGN: Canva trained Craft, its first in-house image model, on human-designed templates instead of photos. Craft makes up to 2K images in 30 aspect ratios with legible text. Canva Magic, its image-to-design tool, can split them into editable layers and every user gets it. Nice.
AI SHOPPING: Landed on an AI assistant’s shortlist? Your reviews still close the sale. Only 5% of shoppers buy a recommended product on the spot, per Feedvisor. After an AI suggestion, 33% check reviews, while just 14.8% visit the brand’s site, PartnerCentric found. Those assistants read your reviews too, so good ones help you get recommended.
OPENAI: Like Jurassic Park’s raptors, AI models test the fences. On September 20, an internal OpenAI model exploited a DNS misconfiguration to get limited internet access mid-training. OpenAI paused its top models for 11 days and set its own AI agents on the defenses. They found several ways out. Now those models train offline.
CHATGPT ADS: Amazon might be your new way into ChatGPT ads. Juozas Kaziukėnas spotted some whose clicks route through Amazon’s click-tracking server, a sign that Amazon DSP, the company’s platform for buying ads across the web, serves them. Details are thin, and he hasn’t seen other advertisers use it yet. Buying through DSP? Ask your rep.
ANTHROPIC: Even AI companies can’t prompt their way to brand awareness. Anthropic’s IPO filing, the first public look at a major AI lab’s ad spend, shows $33 million in 2024 and $147 million in 2025. MediaPost projects over $655 million this year, calling AI brands a possible multibillion-dollar ad category. Selling ads? Pitch them.
GOOGLE: Searched your own brand lately? Google’s AI probably spoke first. AI Overviews, the summaries above Google’s regular results, now appear on over 80% of searches, up from 50% in June 2025 and about 15% a year earlier. SE Ranking found them on 82.5% of US brand searches. Check what it says about you.
Get the next AI tactic in your inbox.
Ready-to-apply AI tactics delivered every Saturday to help you get wins on Monday. Consumed in 7 minutes or less.
