A million tokens in context means you can fit an entire codebase in a single prompt now.
The question nobody's answering yet: can the model actually do useful work with something 500,000 tokens back?
Context window ≠ comprehension window.
Bootstrapping a global dataset is less about scraping…
and more about building the right pipeline.
For Mossie we built a serverless ETL pipeline where every bird runs through its own workflow.
Parallel, observable, and weirdly full of chickens and riflemen.
Your engineering team is probably your most expensive investment.
Your responsibility as a founder: Make sure they’re building the right thing.
That starts with talking to users, not brainstorming features.
True or false?
Someone asked:
“Can we add all the birds?”
Turns out that means 11,089 species, with images, sounds, and metadata.
So we built a GenAI pipeline to bootstrap the database.
It involved hallucinations, broken taxonomy, and a scary OpenAI invoice.
Instead of telling an AI agent to "write better code," give it a CI pipeline that fails on bad code.
PHPStan + PHPInsights as quality gates. Measurable. Objective. No subjective feedback loops. Agents should fail on metrics, not vibes.
"Remote-first decoupled teams from offices. AI decouples them from hardware."
The flattering version of external consulting is that outsiders are more courageous than internal teams. That we see things more clearly, speak more freely, act more decisively.
The honest version is simpler. We're temporary.
“Wouldn’t it be nice if…”
That sentence has killed more engineering focus than most bugs ever did.
Good founders filter ideas.
Great founders protect their team’s focus.
Everyone is chasing the “perfect AI prompt”.
Meanwhile, the real leverage is knowing what problem you’re trying to solve.
Prompts don’t replace thinking.
In a world where AI makes knowledge cheaper by the month, the part that stays expensive is the willingness to act on what you know.
We tried to add every bird in the world to a birding app.
11,089 species.
Photos. Sounds. Metadata. Icons.
The solution: a GenAI-powered pipeline that bootstraps the whole database.
And yes… it got weird.
What if your team's next bonding activity was actually fun? We've been building Take Five: a quiz game with duels, sabotages, winning streaks, and way too many "wait, you didn't know that about me?" moments. Waitlist is open: take-five.madewithlovehq.com
I built a quiz game at @madewithlove.com for remote teams: think pub quiz meets Football Manager. Social questions about colleagues, seasonal play, duels, hall of fames, streaks, sabotages,...
Free (for now) for early companies (just give us feedback).
Waiting list: take-five.madewithlovehq.com
The Jury members have cast their final votes after the pitching round, and the winner is... 🥁 Eply ! Congratulations! They will receive a full technical audit to validate architecture, increase velocity, and align product & tech strategy for the next 18–24 months. And tons of insights and tips.
Congratulations, Enkryptify, Eply and WeGroup: the Jury of the Jump Start Competition has nominated you as finalists to go through to the final pitching round! We look forward to your video pitches! Decision Day is 6 February.
"A comet is headed straight for the planet of the manual coders. By 2030, nobody will be writing code anymore." #software #engineering #founders #buildinpublic #SaaS #AI
LAST CALL to submit your SaaS and/or AI startup for our Jump Start Competition 🦘 and seize the opportunity to win our expert technical support to help you reach your next milestone! Submissions are closing on Tuesday evening. #SaaS #AI #startups #startup #founders #investors #VC #seed
The first future-shaping startups submitted their 💼 - don't miss out on YOUR chance to give your SaaS or AI 🚀 product a ⚡ with +15.000 euro worth of technical expertise and coaching by our resident CTOs!
#startups #startup #SaaS #AI #founders #founder #buildinpublic
The Jump Start Competition 🦘 is now officially OPEN! Enter 📥 your SaaS or AI company today, and tell our Jury 👨⚖️👩⚖️ why they should choose YOU to give your business the technical edge 🗡️ that you need to scale 📈 successfully and reach your next milestone! 🏁
#startups #founders #AI #SaaS #buildinpublic
SaaS/AI founder? Pre-register for a chance to WIN (value >15.000 euro!) a full technical audit and personal coaching. Get an independent review of your product, architecture, and engineering operations by our resident CTOs. And benefit from CTO mentoring to help you scale faster. bit.ly/mwljumpstart
Wouldn't it be nice if someone could give your SaaS or AI company's tech team a JUMP 🦘 START and help you scale faster? Wouldn't it be even nicer if you could win 🏆 that FOR FREE 🎁 ?
Check it out and leave your details to get notified when it kicks 🦶 off in January: bit.ly/mwljumpstart
"Performance bugs erode trust quietly until users explode. Three pragmatic steps help you catch slowdowns early: explore real bottlenecks with Sentry, test with production-sized data, and add lightweight API load tests."
"Cloud outages arrive with the same elegance as a falling wardrobe. One moment everything is humming along. The next, half your platform is unreachable, support inboxes are filling up, and someone is muttering about DNS again."
"Of course your product is AI-based. There is no need mention this anymore."
