#random - January 2026

The OSA Community is excited to host the Open Lakehouse and AI events in New York and Chicago!

Registration links:
🗽New York meetup: https://luma.com/dvopjirh
Featuring talks from Altinity, Dremio, CelerData, and Grafana

🏙️ Chicago meetup: https://luma.com/699ffm7v
Featuring talks from Altinity, Fivetran, and CelerData

I’m looking for a California-based (LA, SD, or SF) technical co-founder to explore a DevSecOps / DevOps startup venture. I bring a GTM/Product Research focus and am seeking a partner with a complementary technical skillset to build and scale. If you’re local and interested in shifting security left, please DM me to discuss the vision.

Hi all, quick ask. I’m trying to find the right professional channels to share an upcoming technical webinar (no product pitch). It’s about practical multi-cloud security implementation using native controls (AWS/Azure/GCP) and how teams scale it across accounts, orgs, and environments.
If you know relevant Slack channels, LinkedIn groups, or newsletters where this fits, I’d appreciate pointers.

If anyone is struggling with test pollution in ephemeral environments, we wrote up a guide on how to handle the data layer.
It covers how to trigger ephemeral DB branches (using Neon) that match the lifecycle of your K8s sandboxes. The main win is using copy-on-write to get instant branches so you aren't waiting on database provisioning for every PR.
Here’s the tutorial if you’re interested: https://www.signadot.com/docs/integrations/databases/neon

Hey everyone, I’m working on a tool to help freelancers and small agencies respond to RFPs quickly instead of manual copy pasting and coordinating with different SMEs for review.

The specific problem I’m trying to solve is, first, helping teams quickly decide if the RFP is worth pursuing before actually working on it. Second, to reduce users’ time on manually finding questions and copy-pasting answers from previous RFPs and then getting those reviewed by SMEs in multiple email threads.

It’s still pretty early and I’m still figuring out what’s the best way possible to reduce time and effort.

If anyone has dealt with RFPs before, I’d really appreciate any inputs like what part of the RFP response process is most painful and consumes most time.
Thank you

Hey hey!

I’m pleased to share an app I’ve been working on!

Before Leaving is designed to help users stay organized and ensure nothing important is forgotten before heading out. I would greatly appreciate your feedback and thoughts.

Before Leaving
https://lnkd.in/d5ASc94x

Hey folks — I’m Satish from PrArySoft, building IncidentIQ, a Slack-native incident capture + summary tool.
It does /incident start + /incident end, then posts a clean summary, key events, action items, and a markdown export.
I’m recruiting 3–5 early beta teams (Slack-first incident response) and looking for honest feedback.
If anyone wants to test it during real incidents, DM me and I’ll share the install link + 1-page quick start.

Ex:

Hey folks, Blast is hosting a live webinar with Rapyd’s CISO on why preemptive cloud defense is becoming the standard in 2026.
With non-human identities and IAM operating at machine speed, detection and response alone can’t close the risk window fast enough. Prevention is often the fastest mitigation.
👉️ Register here: Click here

👀 Check our new blog post: "LLM Agents in Clojure" — practical examples, stubs, and traceable runs you can try in the REPL.
🔗 15‑min read → https://www.freshcodeit.com/blog/llm-agents-in-clojure

Interesting read on the shift from manual code review to behavioral verification.
The argument is that checking syntax doesn't scale with AI agents, and is becoming irrelevant anyway, so the deploy preview becomes the primary gate.
Curious how others are thinking about this transition? Are you seeing a move away from line-by-line review yet?
https://thenewstack.io/traditional-code-review-is-dead-what-comes-next/

Hello all! We are experimenting on a product from our daily struggles and try to gather some feedback / understanding on what is everyone going through. It would mean a lot if any of you can hep us with this 10-15 mins questionnaire 🙏

https://docs.google.com/forms/d/e/1FAIpQLSdSOnulTc4NDR_GfAdv2EdZ7z0Bjh58yUz2d0NGYtHyMjukvg/viewform?pli=1&authuser=0

Hi everyone, I’ve recently started writing blogs on infrastructure-related topics. So far, I’ve written three blogs:

Running GPUs in Kubernetes: From Setup to Scheduling and Sharing
https://medium.com/@pabhi18/running-gpus-in-kubernetes-from-setup-to-scheduling-and-sharing-b6b7ebd10d4e

Inference Engine: A Simple Explanation
https://medium.com/@pabhi18/inference-engine-a-simple-explanation-80d319a492be

Running LLM Inference on GPUs with Kubernetes
https://medium.com/@pabhi18/running-llm-inference-on-gpus-with-kubernetes-41ce0d0c5594

The third blog is based on a project I recently built and I’d love to hear your opinions and feedback on my blogs.

I just shipped a new MCP server to align with a SEP we opened called Hotcross. It persists static code graphs + other agent context into a local chDB database, so agents can query durable project knowledge with SQL.

Highlights:

- 💾 Persistent storage in portable .db files
- 🔍️ SQL queries over stored context
- 🧠 Tree-sitter code intelligence (symbols, deps, references)
- 📝 Session memory (LLM chat history)
- ✅️ Todo/project tracking

Repo + quickstart: https://github.com/atelierlogos/hotcross

Would love your thoughts if this is in your wheelhouse.

Hi guys,

If anyone is having a hard time with web app security, need additional firewall, ddos mitigation (l3,4,7), WAF or CDN

Please get in touch - I will be your incident response coordinator and can get you migrated to security in any tech stack, managed or un managed DNS with Vecurity, an up coming leader in web security

Otherwise - https://www.vecurity.net/

Thanks

Check out this piece on The New Stack about using environment virtualization to give coding agents runtime context so they can verify what they write.
Might be worth a read for anyone looking to bring more autonomy into their agentic workflow:
https://thenewstack.io/enabling-autonomous-agents-with-environment-virtualization/

exciting!
https://oban.pro/articles/introducing-oban-python

For anyone with a python stack and you'd like to simplify some infra

Wrote a blog explaining how inference engines work under the hood and how GPU metrics change with them. If you’re interested, check it out here

https://medium.com/@pabhi18/how-llm-inference-works-under-the-hood-prompt-processing-and-gpu-behavior-0b69a222a0b6

Interesting article on Ramp’s Inspect platform and how closed-loop agents are shaping the SDLC.
Worth a read for anyone who is thinking about how to turn coding agents into actual developer velocity.
Read the full breakdown: https://thenewstack.io/ramps-inspect-shows-closed-loop-ai-agents-are-softwares-future/

Most Heroku alternatives get compared by features, but in practice the billing model is what really shapes your experience.
I put together a decision framework based on real production usage and actual costs from Jan 2026, focusing on when Railway, Render, Fly.io, or fixed tiers actually make sense.
https://seenode.com/blog/heroku-alternatives-decision-framework