Naledi Kekana

Creative. Futurist. Advocate

Building and learning in public

[Naledi Kekana @naledicodes]

https://t.co/Xq9xfa9q2G

[Thariq @trq212]

https://t.co/R6exTuF7P8

[Lucas Rizzotto @_LucasRizzotto]

I'm building cinematic games for movie theaters!

We're showing our first this week in one of the biggest film festivals in the world!

[Naledi Kekana @naledicodes]

https://t.co/bc6ghvoTN7

[TensorTonic @TensorTonic]

We're hiring for -

- AI engineering intern
- GTM/Growth intern

Remote. Growing at insane speed. 40K users. High agency.

If you're cracked and want to work at early stage startup, then consider applying. More details below 👇

[Google for Startups @GoogleStartups]

Unlock the power of autonomous workflows with Startup School: Agentic AI.

Join @GoogleCloud experts to discover how to build, deploy, and scale sophisticated #AIagents that drive real business impact. Ready to innovate? Register here ➡️ https://t.co/QzUDNTjYzt

[Alexander Morales-Panitz @moralespanitz]

read colleague.skill this morning and honestly it made me think we've been building memory systems completely wrong...

the problem
-----
every memory system today is a solo instrument. one agent, one session, one recall. you store facts, you retrieve facts. but when someone leaves your team you don't lose their files, you lose the reasoning behind the files. why that decision was made, what tradeoff they were optimizing for. current systems don't capture that at all

what they actually built
-----
you feed it someone's slack history, docs, emails and it distills two things separately. the work skill which is how they make decisions and their workflows, and the persona which is just how they communicate. those are separate layers and that composability is genuinely smart. most systems bake both together and make both worse

what hit me the most
-----
this is the first clean formalization i've seen of what i'd call collaborative context as infrastructure. knowledge that doesn't just sit in a store but travels between agents and sessions with the reasoning intact. that's not a better memory system, that's a different category entirely

the limitation that nobody is talking about
-----
their evolution mechanism is input triggered. new data arrives, skill regenerates. that's fine but it's reactive. what you actually want is performance triggered evolution. the system detects its own behavior is degrading and updates without waiting for someone to feed it new docs. that's the hard problem and it's still unsolved

the opportunity
-----
whoever builds the performance triggered version first, with local first privacy guarantees, basically owns the next layer of how agents collaborate and share knowledge

we are genuinely early on this one

[Naledi Kekana @naledicodes]

https://t.co/kmBPKHZpWq

[GitHub Projects Community @GithubProjects]

No more writing OAuth flows, token refresh logic, or rate-limit handling from scratch.

- Managed auth, proxy, and functions for 800+ APIs
- Write integration logic as TypeScript functions or generate with AI
- Built-in token storage, retries, and multi-tenant connection management
- Production runtime with observability and scaling

[cat @_catwu]

Excited to share our most powerful new Claude Code feature: dynamic workflows!

Mention "workflow" in a prompt and Claude will dynamically create an orchestration plan that it strictly follows, allowing you to confidently trust that every stage happens in the right order even across 100s of agents.

[Kris Lachance @KrisLachance]

We’re hiring a Senior Full Stack Engineer at
@Basedash!

We’re a small team building the leading AI-native Business Intelligence platform.

BI is a massive market entering a major shift with new ways of storing and computing data, and deeper governance for AI agents operating on company data.

We’re looking for someone with strong full-stack fundamentals, real product sense, and genuine excitement about building with AI at the core of the product from day one.

[Mesh | We Are Hiring! @meshpay]

👥 Meet the team helping Mesh scale!

As Mesh continues to grow across products, markets, and time zones, so does the team behind the team: our lovely HR department.

Want to join us? We're hiring → https://t.co/T0Zy3nXqE4 🚀

[Lidia • Technologist 👩🏽‍💻 @poderosacoding]

Hi All!

I’m looking for recommendations for a trustworthy and reliable immigration attorney or nonprofit legal organization experienced with asylum cases in NYC/NJ.

This is for someone with an important immigration appointment coming up on June 10, so I’m especially looking for attorneys or organizations that are responsive, compassionate, honest, and experienced with asylum/immigration court matters.

If you’ve had a positive personal experience or know someone reputable, please DM me or comment below. Thank you so much.

[Lakshya Shukla @DevByLakshya]

🚨 AI companies are now hiring a new type of engineer:
FDE = Forward Deployed Engineer

Companies like anthropic, openai, and others are racing to hire them.

What they do:
- Work directly with enterprise clients.
-Build custom AI workflows/apps.
-Fine-tune AI integrations.
-Ship production AI fast.

This role is exploding because companies don’t just want AI models anymore they want engineers who can deploy AI into real businesses.

[Charles 🎉 Frye @charles_irl]

Writing fast, correct kernels is very hard! For both speed and logic, verification is surprisingly tricky.

Very cool work, and writeup, from @_doubleAI_

https://t.co/vuyVEtpUmN

[ben hylak @benhylak]

this saturday, @OpenAI is throwing an autoresearch hackathon with @raindrop_ai and @modal

come learn how to build systems (agents, models, etc) that improve themselves

link below!

[Naledi Kekana @naledicodes]

https://t.co/rSTP67FM4s via @YouTube

[Naledi Kekana @naledicodes]

Hi! I got a personal trainer covered by my health insurance from Kickoff and thought you might like it as well! You can check it out here: https://t.co/hebvRLJLy9

[Naledi Kekana @naledicodes]

@coreyhainesco These are so cool ❤️ She should do a talk on them.