Natik Gadzhi

"We have generally been finding out in the last few days"

"You can't lead a horse to look at its teeth"

I'll be in #Minneapolis for the next 10 days! What's good? Recommend me #record stores, #books stores, #coffee shops, and quiet parks. I'll need space to recover from humans. There will be a lot of humans.

Where's my anxious #ADHD brethren at? How's hydroxyzine long-term for you? Not feeling like breaking stuff for a change is very nice. Does that last?

Engineering leadership: sometimes its performance management, and sometimes its _performance_ management 🤡🫡

@kepanoI've used@obsidian.md for years on Catalyst license, but finally subscribed to sync. First, because I needed to sync up vaults between personal and work laptop, but even more so because I wanted to finally support @stephango.com and the team's work.

file over app
function over form

🖤

This Microsoft-backed AI startup just collapsed after faking its AI services with 700 real engineers In a story that sounds like it's out of a sitcom, "AI" startup Builder.ai, once valued at $1.5 billion and publicly backed by Microsoft, has just collapsed after it was discovered its AI was actually just 700 Indian engineers manually working on tickets.

https://www.windowscentral.com/microsoft/builder-ai-collapse-microsoft-backed-fake-ai-services


Arguably that perhaps was cheaper than inference at openai, I wouldn't be surprised if the board was happy with the strategy.

Senior Software Engineer - Inference Apps In 2012, Lambda started with a crew of AI engineers publishing research at top machine-learning conferences. We began as an AI company built by AI engineers. That hasn't changed. Today, we're on a mission to be the world's top AI computing platform. We equip engineers with the tools to deploy AI that is fast, secure, affordable, and built to scale. Whether they need powerhouse GPU hardware on-site or the flexibility of cloud-based solutions, we've got the horsepower to make it happen. Lambda’s AI Cloud has been adopted by the world’s leading companies and research institutions including Anyscale, Rakuten, The AI Institute, and multiple enterprises with over a trillion dollars of market capitalization. Our goal is to make computation as effortless and ubiquitous as electricity. If you'd like to build the world's best deep learning cloud, join us.  Note: This position requires presence in our San Francisco office location 4 days per week; Lambda’s designated work from home day is currently Tuesday. What You’ll Do - Work on Lambda’s Inference-based apps, from their core components (scalable distributed backend systems, tool calling, multi-tenant data integrations, semantic layer, etc) to user-facing web and platform native applications. - Build systems that make bleeding edge LLM application features available and consistent across the newest open source models, and help define standards around them. Including accessing user’s data, MCP integrations, tool calling, etc. - Own the full feature lifecycle from a hypothesis to fast prototyping, shipping, evaluating feature performance, and iterating on it. - Work closely with Lambda Inference, Core Apps, Platform, and Special Projects teams, other large players in the industry, and the open source community. - Make the newest, highest leverage and most impactful and useful features of language models available to more folks, in a privacy-first, open source-first environment. Our work will help independent researchers get their models into the market. - Help build the newest, ambitious line of business at Lambda. Own the outcomes for the features you will work on, and work directly with stakeholders across the company, as well as our partners outside of the company. You - 10+ years of experience building software products. You’re already a senior, experienced software engineer. - 1 year or more of experience working on LLM-based applications, including hybrid search systems, user data integrations, evals, and tracing. - 5+ years of experience with Python and/or Typescript. This role will focus mainly on the backend first, however, we’re a small team, and you will be expected to write Typescript in our (modern) React and Svelte codebases, as well as prototype frontends for your experimental apps. - You’re resourceful and you take ownership. You’re driven by outcomes, not just output of your work. - You have an insatiable desire to ship. - You take a user-centric approach to building products and developer tools. - A passion for continuous learning and improvement (both at a personal and team level). - You enjoy building leverage through abstractions and tooling that multiply output. - Ability to influence without authority and work with engineers junior and senior, experience working with operations, platforms, and infrastructure teams. - Excellent written and verbal communication skills. You communicate proactively, you worked with hybrid and remote-first async teams before. Nice to Have - Experience building OAuth integrations. - Experience implementing data pipelines. - Extensive experience building with language models. - Experience leading a team in a tech lead role previously is a strong plus. - Experience starting a company is a strong plus. Salary Range Information Based on market data and other factors, the annual salary range for this position is $255,000 - $405,000. However, a salary higher or lower than this range may be appropriate for a candidate whose qualifications differ meaningfully from those listed in the job description. About Lambda - Founded in 2012, ~350 employees (2024) and growing fast - We offer generous cash & equity compensation - Our investors include Andra Capital, SGW, Andrej Karpathy, ARK Invest, Fincadia Advisors, G Squared, In-Q-Tel (IQT), KHK & Partners, NVIDIA, Pegatron, Supermicro, Wistron, Wiwynn, US Innovative Technology, Gradient Ventures, Mercato Partners, SVB, 1517, Crescent Cove. - We are experiencing extremely high demand for our systems, with quarter over quarter, year over year profitability - Our research papers https://lambdalabs.com/research have been accepted into top machine learning and graphics conferences, including NeurIPS, ICCV, SIGGRAPH, and TOG - Health, dental, and vision coverage for you and your dependents - Commuter/Work from home stipends for select roles - 401k Plan with 2% company match (USA employees) - Flexible Paid Time Off Plan that we all actually use A Final Note: You do not need to match all of the listed expectations to apply for this position. We are committed to building a team with a variety of backgrounds, experiences, and skills. Equal Opportunity Employer Lambda is an Equal Opportunity employer. Applicants are considered without regard to race, color, religion, creed, national origin, age, sex, gender, marital status, sexual orientation and identity, genetic information, veteran status, citizenship, or any other factors prohibited by local, state, or federal law.

My team at Lambda is hiring! We're building LLM-based app toolkits and model-agnostic tool calling and other goodies so folks can use them on top of open source models.

SF only. Hybrid, 4 days in office. If we worked together before, reach out!

https://jobs.ashbyhq.com/Lambda/e325f09b-992e-4c2d-96

Changelog | Cursor - The AI Code Editor New updates and improvements.

The one-click MCP setup with oauth is the story here. https://www.cursor.com/en/changelog?utm_source=alphasignal

This is really a big deal, and I think it'll significantly increase MCP adoption.

PSA: Say thank you to folks in your service and ops teams. Someone helped you onboard and is responsive in your it-support tickets? Someone who helped you get your apps up and running in the k8s cluster?

Say thank you. Show folks you're grateful and appreciate their work. Goes a long way.

Book cover: The Friend by Sigrid Nunez

This was really really good. CW: suicide, grief.

On the off chance that you did not think that Andrew Huberman was a complete dipshit as well as as a hopelessly self-indulgent blatherer, boy do I have a podcast for you.

We're still dialing in even the basic table stakes features (mobile apps, web search quality, file uploads, data integrations), but it'll catch up soon.

If you're down to poke holes in it and tell me what would stop you from replacing Claude in your life with it — rip it to pieces, I'm listening.

And I shipped it to Lambda Chat last night! Lambda.chat is transitioning from a demo stand to a product we will continually invest in.

It's here to stay. It will become better and better. You can sign up and use it with modern, cutting edge open models today.

https://lambda.chat

We already have it live on Lambda Inference API https://lambda.ai/inference-models/deepseek-r1-0528

So far, smooth sailing. You can also use it on Lambda via #openrouter with solid uptime and throughput: https://openrouter.ai/deepseek/deepseek-r1-0528

Open models are _really_ catching up. The R1 update from Wednesday is close to o3 proper (not mini!?) and gemini 2.5 pro (!?).

https://huggingface.co/deepseek-ai/DeepSeek-R1-0528?utm_source=alphasignal

Subtweeting is cool, but have you tried subwebsiting? 🙃

All cloud work and no cloud play makes Natik a dull boy

Horizons into battlegrounds

And one of the experiences that helps folks bond most is facing adversity together, and supporting each other through it. Sometimes it's outside competition. I have a feeling it's not uncommon to have healthy discontent with some things internal. Like a circus-grade product process.

It's not a _company_ that made that best job you've ever had. It's the _people_ and the shared _experiences_.

Which is to say, I'm moving back to the bay 💙

If you're out of ideas for Mother's Day gifts, you can always switch to @kagi.com Family plan and gift your partner the bliss of damn good web search.

Have you considered the ferality of the vibes today?

What makes a genuinely good and useful #llm-based app, and a shitty one? What makes one chat better than another? What features and characteristics are important?

Alright I wrote a thing on how to get career matters into your own hands and hopefully not fuck things up. Beta testing here. What have I missed that worked for you well?

https://respawn.io/posts/how-to-get-promoted

I have 5 one month subscriptions to The Engineering Manager https://theengineeringmanager.substack.com/ by James Stainer to give away. If you're early in your eng leadership journey, his writing is a great resource to learn from.


Hit me up if you want one!

For-profit venture-backed companies making their source available under GPL is a spit in the face, don't @

SMH Zed.

Counterintuitive, but it's after a good workout that I have enough energy to take these two for a walk in the afternoon.

Just me, or using hashtags to farm engagement makes everyone feel dirty?