41, engineer, founder, nocodo: Lovable for AI agents
Joined March 2008
- Tweets 2,358
- Following 552
- Followers 881
- Likes 2,072
144 Photos and videos
Pinned Tweet
15 Oct 2025
Sometimes it takes years for an idea to work out.
I own my product's domain since 2013! Chased and failed multiple times. It was beyond what a solo founder could do.
Then came LLMs, they got better at code. I restarted my product. Fresh perspective and a decade of attempts.
8
594
Jun 11
When everyone is tokenmaxxing, I am building a coding agent with an opinionated tech stack:
- CRUD and RESTful API with Rust Actix Web
- SQLite
- Users, teams, profiles
- RBAC with scopes
- Mobile/web apps with SolidJS
- Admin app
- Hosting backups
github.com/brainless/nocodo
37
May 26
Is it possible to build a coding agent around LLMs that are less than 10B parameters?
What about less than 1B parameters?
Is that crazy?
20
May 19
Coding agents for most humans will basically be about an opinionated tech stack, deployment, tooling and adding custom business logic in API handlers.
Frameworks that can evolve with AI agents...
What am I missing?
31
May 18
Coding agents are a big part of LLM/AI companies revenues at the moment. But coding is much more structured as a problem.
Sure there are many languages but you can build highly constrained agents that only work with Typescript. Or add another language for backend...
1
20
May 18
You can build a better quality coding agent that takes a user's workflow or business process description through an agile process and deliver full-stack apps using real-world roles, like Project Manager, Product Owner, Engineering Manager, etc...
1
31
May 18
Such a coding agent will run cheaper and produce better results. Build on existing open source stack - frontend/backend frameworks, DB, cache, search, auth and auth - all out of the box and use LLMs to customize them.
Focused agents for different parts of the stack.
28
May 15
There are more LLMs released (open weights or closed) every week than we can build good agents to really utilize them. I mean good harness engineering.
What a time to build harnesses. Agents for the world, agents to empower.
1
15
Apr 10
Who is gonna tell this guy?
Lol these people who cannot face reality. People really are sheep - they need a leader, they like the charm of OpenAI or Anthropic.
When other models are dominating and democratizing, these people will go out of their way to share scam.
Apr 9
all people telling you to switch to this Chinese model as they are "as good as Opus for 1/10 of the price" are basically liars and incompetent
they don't use the tools, they make tweets for hype, because using this model for 2 minutes makes you realize how dumb they are
what for? seriously
65
Mar 31
Multiple entity extraction (financial transactions, orders, locations, events, persons, orgs...) now works with @Alibaba_Qwen Qwen 3.5 0.8b model - should work on any small laptop
Tested on M4 Mac Mini 16 GB @UnslothAI model on llama.cpp)
github.com/brainless/dwata
1
72
Mar 27
I needed GliNER2 for a Rust based desktop app. So I create a quick port of GliNER2 to Rust/Candle:
github.com/brainless/gliner2…
2
68
Introducing GLM-4.7-Flash: Your local coding and agentic assistant.
Setting a new standard for the 30B class, GLM-4.7-Flash balances high performance with efficiency, making it the perfect lightweight deployment option. Beyond coding, it is also recommended for creative writing, translation, long-context tasks, and roleplay.
Weights: huggingface.co/zai-org/GLM-4…
API: docs.z.ai/guides/overview/pr…
- GLM-4.7-Flash: Free (1 concurrency)
- GLM-4.7-FlashX: High-Speed and Affordable
381
713
5,343
1,651,164
Jan 11
The Model Wars is upon us 🍿
Jan 9
Scoop: xAI staff had been using Anthropic’s models internally through Cursor—until Anthropic cut off the startup’s access this week.
Here’s what xAI cofounder Tony Wu sent to staff on Wednesday according to a copy of the internal Slack message I viewed:
“Hi team, I believe many of you have already discovered that anthropic models are not responding on cursor. According to cursor this is a new policy anthropic is enforcing for all its major competitors.
This is a both bad and good news. We will get a hit on productivity, but it rly pushes us to develop our own coding product / models. We're at a time in which Al is now a critical technology for our own productivity. This coming year is rly going to be wildly exciting for all of us.
The team is rapidly developing our own models / product. We will have something to share with everyone soon. In the meantime, you may still try all different kinds of models in grok build.”
Anthropic declined to comment. A Cursor spokesperson directed me to Anthropic for comment. xAI did not respond to request for comment.
1
44
Sumit Datta retweeted
26 Dec 2025
People ask what faster AI actually means for them. Here’s a concrete example. In the video, GLM-4.6 on @cerebras builds Space Invaders in ~15% of the time it takes Claude Sonnet 4.5 Thinking.
The point isn’t speed for its own sake.
When latency drops far enough, new classes of workflows become possible.
This is like Netflix.
Netflix didn’t build a $400 billion business because it mailed DVDs faster. It's entire business changed. It became a movie studio. This transformation was made possible when streaming became fast and reliable enough that people stopped thinking about buffering, downloads, and storage altogether.
At that point, behavior changed: people browsed, clicked, abandoned, re-tried - without friction.
Fast inference does the same thing for AI.
When responses are slow, you design around waiting.
25
20
243
33,003
25 Dec 2025
Nvidia is upset! Nvidia's interests in Groq shows what they see as future threat - open source models powering agents.
Groq is not being acquired, not yet. But let's hope Cerebras and others with inference silicon keep marching ahead. Customers are winning & Nvidia isn't happy
1
114
25 Dec 2025
Apple may take the lead on off the shelf inference and then Nvidia will have a tough time.
Google TPUs are a bigger threat in the enterprise
46
GLM-4.7 is here!
GLM-4.7 surpasses GLM-4.6 with substantial improvements in coding, complex reasoning, and tool usage, setting new open-source SOTA standards. It also boosts performance in chat, creative writing, and role-play scenarios.
Default Model for Coding Plan: z.ai/subscribe
Try it now: chat.z.ai
Weights: huggingface.co/zai-org/GLM-4…
Tech Blog: z.ai/blog/glm-4.7
234
520
3,398
1,422,895
20 Dec 2025
How I structure my projects for easier build with coding agents, Claude Code, opencode, Cursor...
youtu.be/44-xXzLclYE
Really basic, for full-stack apps
- No CLAUDE.md anymore
- Naming - process & conventions
- Break into modules
- Extensive unit tests
1
1
76
18 Dec 2025
Not to cause alarm, just sharing what I see as a pattern...
LLMs will wash away a lot of the economic values we have built
And we cannot stop it. I know it sounds sad but it is what is happening
A huge amount of the world's knowledge was taken to build the LLMs
1
1
30
18 Dec 2025
Then we distill into smaller, faster models which are great at certain things or even competitive with generic tasks but cheaper
Look at the Chinese models for example
We keep adding specific knowledge to the top tier models. They hire lawyers, engineers and so on to train AI
1
29
18 Dec 2025
Just like recent "AI enabled search" on Google is taking traffic away from websites, the same will happen to knowledge economy
And no one has a clue how we will migrate all the people who will not be able to compete with machines
22