The move by Klarna to shut down Salesforce and Workday in favor of a custom AI-driven tech stack is a fascinating signal of where enterprise technology might be heading. This shift reflects a broader trend of companies rethinking the SaaS model and leveraging AI to build more tailored, efficient, and cost-effective solutions.
Here’s why this is significant:
1. Customization and Flexibility:
AI engineers can now create applications that are fine-tuned to an organization’s exact needs, as opposed to relying on the one-size-fits-all solutions provided by traditional SaaS platforms. The power of AI allows businesses to build systems that respond dynamically to specific workflows, which is more efficient than generic tools like Salesforce or Workday.
Klarna’s decision could set a precedent where businesses move away from SaaS products with pre-built constraints and opt for AI-generated applications that are fully customized to their processes.
2. Cost Efficiency:
SaaS platforms like Salesforce and Workday can be expensive to maintain, both in terms of subscription fees and the complexity involved in customizing and integrating them into a company’s operations.
AI-driven solutions are emerging as a way to cut these costs dramatically. Klarna’s custom stack could be 10x cheaper to run, as AI allows for highly automated development, maintenance, and scaling processes. You can cut out layers of unnecessary features and create lean, purpose-built tools.
3. AI Engineers’ Impact:
As more AI engineers become proficient, they’ll have the ability to replace traditional SaaS platforms with bespoke applications that are more efficient and faster to develop. Instead of relying on SaaS ecosystems, AI models can create, iterate, and optimize internal systems on demand.
The long-term effect will likely be a growing shift away from subscription-heavy SaaS models towards more agile, AI-powered solutions that offer greater control and cost benefits to businesses.
4. Risks and Challenges:
While this approach is highly appealing for tech-forward companies like Klarna, there are challenges related to the robustness, scalability, and long-term maintenance of AI-generated systems. SaaS providers like Salesforce have years of industry-specific features and support that a bespoke AI system might lack at the start.
Transitioning to AI-powered systems also requires deep in-house expertise, which might not be available for every organization. It will likely take time before this approach becomes a widespread trend beyond large, tech-savvy enterprises.
5. Future of SaaS:
This shift towards AI-driven solutions could pressure traditional SaaS vendors to rethink their offerings. They may need to pivot towards providing more flexible, AI-enabled options or risk losing enterprise customers who can develop better, cheaper alternatives in-house.
Klarna’s move may accelerate the shift towards a future where businesses prioritize owning their technology stacks, designed and managed with AI, over renting SaaS services.
This development reflects the broader theme of AI democratizing software development, reducing reliance on massive SaaS companies, and giving businesses more autonomy and agility over their tech infrastructure. Klarna’s bold move may inspire others to follow, leading to a fundamental reshaping of the enterprise software landscape.
cc @AdrianDGrant
@ShopifySupport apparently someone tried to fraudulently redirect my domain. I have been trying to get help for a week now, every advisor says the “specialists are working g on it” but no other info. This is the wordy customer service I have had with you, what is going on?
Bryan Johnson is spending $2,000,000/year to live until 120.
So I spent a full day researching his entire longevity blueprint.
Here’s a full breakdown of his routine & how anyone can use it:
(without spending $2M/year)
Day 15 of 95 #javascript: understand break and continue statements and the difference adding a label has on where each jumps to. Review of return statement. Much deeper understanding of try/catch/finally and using it w while loop to replicate a for loop.
Day 14 of 95 #javascript: reviewing Loops. Learned the difference between for/of and for/in (iterable vs non-iterable). Learned about set and map classes.
Day 13 of 95 #javascript: reviewing statements: expression, compound and empty. Reviewing Conditionals: if and else if. Learned about switch, that was completely new to me. Now I’m in new territory. 16% of the way through the book.
Day 12 of 95 #javascript: finished chapter 4, revisited conditional operator, learned First-Defined (??), learned nuances to delete, void, and comma operators. Getting back into reading more frequently.
Day 11 of 95 #javascript: review of logical and assignment expressions. Really dug into when the value of an assignment expression is used as part of a larger expression. Ignored eval() section. Falling behind but I’m determined to finish all 95 days. 💪
Day 10 of 95 #javascript: relational expressions, equality and inequality operators review,comparison operators, and I learned about in and instance of operators for the first time ✌️
Day 9 of 95 #javascript: review function definition expressions and property access expressions. I now understand conditional property access. Review of invocation and conditional invocation expressions. Operators overview
Day 8 of 95 #javascript: learned Destructuring Assignment syntax. Now I understand the iterable object “…rest”. Review of Primary expressions and object and array initializers.
Day 7 of 95 #javascript: Object-to-string, Object-to-number, and special case operator conversions. prefer-string, prefer-number, and no-preference algorithms. Review of let const and var. I learned that properties created with global var can't be deleted with delete operator.
Day 6 of 95 #javascript: Reviewing immutable primitive values and mutable object references. Type conversions in depth. I learned a lot, like exactly why an empty array converts to number 0, infinity vs -infinity, why !!x works, more review of radix (base) 🦾
Day 5 of 95 #javascript: getting harder …went deeper into strings, learned what Unicode nfc normalization is, escaping w back ticks, pattern matching, 6 falsy values, null vs undefined review, and Symbols (took time for me to grasp 😅)
Day 4 of 95 #javascript review of types, values, and variables. I learned how to break down base 16, base 8, and base 2 integers. Went through all Math object properties.
Day 3 of 95 #javascript. Review of Lexical structure: comments, literals , Unicode, and semicolons. I learned I could write emojis using Unicode escapes. Also in depth understanding of why semicolons can be omitted in js.