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
๐๐จ๐ฎ๐ซ ๐๐๐ ๐ข๐ฌ ๐ฌ๐ฅ๐จ๐ฐ. ๐๐ก๐๐ญ ๐๐จ ๐ฒ๐จ๐ฎ ๐จ๐ฉ๐ญ๐ข๐ฆ๐ข๐ณ๐ ๐๐ข๐ซ๐ฌ๐ญ?
When API response times start climbing, the first instinct is often to add more resources, introduce caching, or start rewriting code.
But a slow API doesnโt always mean a slow application.
The bottleneck could be hiding almost anywhere:
โ ๐๐ง ๐ฎ๐ง๐จ๐ฉ๐ญ๐ข๐ฆ๐ข๐ณ๐๐ ๐๐๐ญ๐๐๐๐ฌ๐ ๐ช๐ฎ๐๐ซ๐ฒ
โ ๐ ๐ฆ๐ข๐ฌ๐ฌ๐ข๐ง๐ ๐ข๐ง๐๐๐ฑ ๐จ๐ซ ๐+๐ ๐ฉ๐ซ๐จ๐๐ฅ๐๐ฆ
โ ๐๐ฅ๐จ๐๐ค๐ข๐ง๐ ๐/๐ ๐ข๐ง๐ฌ๐ข๐๐ ๐ญ๐ก๐ ๐๐ฉ๐ฉ๐ฅ๐ข๐๐๐ญ๐ข๐จ๐ง
โ ๐๐๐๐ฏ๐ฒ ๐ฌ๐๐ซ๐ข๐๐ฅ๐ข๐ณ๐๐ญ๐ข๐จ๐ง
โ ๐ ๐ฌ๐ฅ๐จ๐ฐ ๐ญ๐ก๐ข๐ซ๐-๐ฉ๐๐ซ๐ญ๐ฒ ๐ฌ๐๐ซ๐ฏ๐ข๐๐
โ ๐๐๐ญ๐ฐ๐จ๐ซ๐ค ๐ฅ๐๐ญ๐๐ง๐๐ฒ ๐จ๐ซ ๐ญ๐ข๐ฆ๐๐จ๐ฎ๐ญ๐ฌ
โ ๐๐ฑ๐ก๐๐ฎ๐ฌ๐ญ๐๐ ๐๐๐ญ๐๐๐๐ฌ๐ ๐๐จ๐ง๐ง๐๐๐ญ๐ข๐จ๐ง ๐ฉ๐จ๐จ๐ฅ๐ฌ
โ ๐๐๐ ๐จ๐ซ ๐ฆ๐๐ฆ๐จ๐ซ๐ฒ ๐ฉ๐ซ๐๐ฌ๐ฌ๐ฎ๐ซ๐
And sometimes adding Redis, more workers, or additional infrastructure only hides the real problem instead of solving it.
Thatโs why I prefer a simple principle: ๐๐๐๐ฌ๐ฎ๐ซ๐ ๐๐ข๐ซ๐ฌ๐ญ. ๐๐ฉ๐ญ๐ข๐ฆ๐ข๐ณ๐ ๐ฌ๐๐๐จ๐ง๐.
In production backend engineering, performance optimization should be driven by evidence, not assumptions.
Iโve summarized the areas I usually think about when investigating API latency in the post below.
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
Title: Digital SAT Prep Video โ Scripted & Animated
Description: Built a full Manim scene script for a Digital SAT explainer video, including a custom color palette, portrait-mode video config for short-form platforms, and timed scene sequencing with synced text animations.
Tools: Manim, Python
๐ PMax vs Search: What Should You Use?
A common question I hear:
โShould I run Performance Max or Search campaigns?โ
The answer depends on your business goal, audience, and campaign data.
๐ Search Ads
Best when you want to target people actively searching for specific products or services.
โ
More control over keywords
โ
Strong search intent
โ
Useful for capturing demand
โ
Easier to understand search-based performance
๐ค Performance Max
Best when you want Google to use automation across multiple Google channels and optimize toward your conversion goals.
โ
Automated bidding & targeting
โ
Access to multiple Google placements
โ
Useful for eCommerce and conversion-focused campaigns
โ
Can help scale when you have enough quality conversion data
๐ก The key isn't choosing one simply because it's popular.
Start with your business objective, conversion tracking, budget, and available dataโthen choose the campaign structure that fits.
Donโt choose the campaign type first. Choose the strategy first.
๐ฉ Need help planning or optimizing your Google Ads campaigns? Letโs connect.
#GoogleAds #PerformanceMax #PMax #SearchAds #GoogleAdvertising #PPC #DigitalMarketing #AdOptimization #PerformanceMarketing
Great breakdown. The key takeaway is spot on: the campaign type should follow the business goal, conversion data, and available budgetโnot the other way around. Strategy first, campaign structure second.
A context window is a limit, not a budget.
That distinction gets lost once windows reach hundreds of thousands of tokens. The cost of a request follows what you actually send, not how much room was available. A bigger window doesn't make an abandoned branch free. It just lets you carry three of them before anything visibly breaks.
In a long working session, most of what piles up isn't the active task. It's decisions that were later overturned, approaches you tried and dropped, the same instruction restated because it wasn't clear the earlier one still applied, and a schema pulled in for one question that is still loaded twenty exchanges later.
All of it gets processed alongside the material that matters. That shows up in the bill, and it can show up in latency and output quality too.
A quick test: if you started this exact task fresh, how much of what's loaded right now would you bring with you?
I wrote up the longer breakdown here: https://medium.com/@nagatomopedro05/the-bigger-context-window-didnt-save-you-any-money-here-s-the-receipt-73f6236ee607
