Aabhas Sharma

Really great to see companies like Valo creating space for the community to come together, learn, and share what actually matters. Bringing admins, architects, and developers together to reflect on new releases, exchange perspectives, and learn from one another is how the ecosystem stays strong and grounded in real-world needs. Appreciate the leadership from Jari Salomaa and the Valo team for investing in community, conversation, and connection. This looks like a fantastic opportunity.

13

1 Comment

Waymo has unlocked a surprisingly specific use case in San Francisco: parents are trusting it to drive their kids to school. They wouldn’t put their kid in an Uber, but they’ll send them solo in a fully autonomous car. That was the moment it really clicked for me. If AI can handle real-time decisions where safety is on the line, it can handle a mortgage file. Document review, condition clearing, data extraction—it’s not even close in terms of complexity. The challenge isn’t whether it’s possible. It’s whether we’re designing systems that make better use of people’s time—because the manual work isn’t just inefficient. It’s holding the entire industry back. https://lnkd.in/gFVA9Yvk #MortgageTech #Automation #AIinLending

12

1 Comment

Marcus built his tech company from a garage startup to a $50 million enterprise. His board celebrated record growth. His employees admired his vision. His investors praised his strategic brilliance. Yet in the quiet moments between meetings, he struggled with something he couldn't discuss with anyone: a persistent sense that he was failing at everything that mattered. His marriage was strained. His sleep was fragmented. His confidence in major decisions had eroded to the point where he'd spend hours second-guessing himself. When his CFO mentioned therapy, Marcus dismissed it immediately. CEOs don't have time for that, he thought. More importantly, what would happen if word got out? The research tells a different story: → 55% of CEOs experienced a mental health issue in the past year (up 24 points from previous year) → 50% report feeling lonely in their roles → 61% believe this isolation negatively impacts their performance → 81% believe organizations view mental health issues as weakness → 70% of C-suite leaders seriously considered quitting for roles that better support their wellbeing Here's what most people miss: the very traits that propel someone to the CEO role—relentless drive, high standards, comfort with risk, the ability to compartmentalize—can become liabilities when psychological strain reaches a critical point. The isolation of leadership creates a perfect storm: immense pressure combined with virtually no safe outlet for processing it. Addressing these challenges doesn't compromise your leadership. It enhances it. https://lnkd.in/gJdyKVJd #CEOMentalHealth #Leadership #ExecutiveWellbeing

1

New blog from the Dropbox Dash team on how we use context engineering for agents, and one lesson stands above the rest: context is the real constraint. When you start giving models the ability to reason and act, not just retrieve, you realize that more isn't always better. More tools, more data, and more context makes the model worse at its job. Here's what we discovered 👇 1. Too many tools = too much noise Each tool definition adds tokens and complexity. As we added more capabilities, the model's reasoning slowed down. It spent more time deciding how to act instead of actually acting. 2. Bigger isn't better Dumping more data into the prompt didn't help. It made the model distracted. Classic context rot. We found that a smaller, more relevant slice of context from a knowledge graph consistently led to clearer reasoning and faster execution. 3. Specialized agents beat general prompts Some tasks like query construction or search, needed deeper focus. We built specialized agents for those. 4. Consolidation unlocks capability We replaced dozens of separate retrieval APIs with one universal search tool. That single interface simplified planning, improved consistency, and freed up valuable context space for actual reasoning. The key takeaway: Building agentic systems at scale isn't about giving models more to think about. It's about giving them just enough of the right things. Read the full post here: https://lnkd.in/gQM7s-q8

94

2 Comments

To date, writing extensions for Envoy has been a laborious process that involves writing C++ and compiling a fully custom build of the entire proxy.  Can we make that easy and make Envoy more approachable by developers?. Can we create an ecosystem of reusable extensions around it - akin to how npm ecosystem fundamentally altered Node.js's trajectory. With that vision, we at Tetrate have been working with Envoy community to first make Envoy easy to extend by any developer. We added dynamic modules a year ago and then added a proper Go and Rust SDK around it. Today, we are making the developer experience much easier and announcing a home for Envoy extensions with an open source extension marketplace around it. Thanks Matt Klein Anna Berenberg for your support and thanks Rohit Agrawal Ben Plotnick Anubhav Aeron for early feedback and trials. You can read more about Built on Envoy here: https://lnkd.in/gbNz9-3v and see a cool demo video from our engineering lead Ignasi Barrera on how to write a simple Go extension and test it locally for Envoy to not allow AI calls to specific models. https://lnkd.in/gcpYJyjD

74

Why is Anthropic hiring engineers like crazy, even though Claude is writing all the code? Boris is the head of Claude Code, this was his response: They still need engineers to prompt, talk to customers, prioritize work, and coordinate with other teams To most people, that sounds more like the scope of a PM But of course, a high-judgment engineer is great at these things as well The lines between job descriptions are blurring

381

32 Comments

Ask AI how OpenClaw can automate workflows in your business, then send me a message and I'll get that all setup for you. Setting up an agentic system can be tricky. Northstar AI Labs has been deep in the weeds, developing AI solutions. If you have heard of AI automation, but you don't want to DIY, a hassle free installation is just 1 message away.

1

I've been discussing this recently with Frances Buontempo in the context of coding assistants. We hear a lot from developers saying that LLM assistants in their fave IDEs are great for generating a lot of the (boring) boilerplate code needed to stand up a non-trivial system. However, most popular IDEs have been able to do exactly this for decades--whether with "wizards", code-block and project templates, or other various forms of meta-programming. The difference is that the "old" way is much less hungry of resources, while simultaneously being much more predictable and reliable at automatically generating the "boring" stuff. Coding assistants are also good at other things, of course, but many of the popular (in my conversations with developers) tasks are those I would much prefer to do myself, such as writing tests.

15

6 Comments

Great interview with Rossella Blatt Vital VP Engineering at Sprout Social, Inc. about what happens after your first #AI feature ships. She digs into the enablement engine your AI team should own so PMs and engineers can build safely at scale with evals guardrails and feedback loops. https://lnkd.in/ebcAvWRm

2

Had a great time hosting and listening to Tamir Dresher talk about scaling and coordinating AI agents. He shared a ton of practical tips for being highly productive with AI agents, including hands-free AI interaction while driving. https://lnkd.in/gtwUN9eG

8

Study Mode is a new feature in ChatGPT designed to help students learn more effectively. It offers personalized explanations, practice questions, and summaries to support focused studying. Just released by OpenAI, it turns ChatGPT into a smart learning assistant.

1

At LinkedIn, we're modernizing hiring data flows with a unified integrations platform. By coexisting BuildIn (partner-owned pull and push integration) and BuildOut (LinkedIn-owned pull and push integration) behind canonical contracts, we reconcile ATS/CRM variability and deliver consistent, near‑real‑time entities to products like Recruiter and Hiring Assistant. Under the hood: Temporal-orchestrated workflows, Kafka streams, record persistence in Espresso, multi mode orchestration and declarative schema/ID mapping enable replayable, bidirectional sync and safe evolution. The payoff: faster partner onboarding, broader data coverage, and higher completeness grounding LinkedIn's Hiring Assistant in trustworthy context and enabling governed actions across systems. A testament to exceptional cross‑functional collaboration, huge kudos to our engineering teams for solving a hard problem and delivering a sophisticated, agentic‑ready, forward‑looking platform! For a more detailed technical roller-coster visit our blog at - https://lnkd.in/UnifiedIntegrationsBlog

51

1 Comment

40% of engineering applicants were fake, and now, hiring teams have a new way to respond. Greenhouse’s integration with CLEAR introduces identity verification as a seamless step in the hiring journey, not a post-offer afterthought. It’s a sign that trust is becoming just as critical as fit. In a world of AI-generated resumes and deepfake interviews, this shift feels timely—and overdue. It won’t solve everything. But it’s a bold move in the right direction. https://lnkd.in/ez9eihjT Greenhouse Software #recruiting #HRTech #WorkTech #IdentityVerification

3

rqlite -- and the way I develop it -- is fundamentally changing. LLMs are in the loop -- from ideation to Pull Request. https://lnkd.in/ejwXGddR

27

1 Comment

There is a new spec manager in town called Conductor, yet another release from Google, which hasn’t stopped shipping for the past few weeks... It is a direct competitor to GitHub’s Spec-Kit and Buildermethods's agent-os, bringing the "Spec-Driven Development" (SDD) philosophy directly into the Google ecosystem. It is deeply integrated with the Gemini CLI and follows the same rigorous standards we've seen in other top-tier agentic tools. Conductor uses a structured workflow to ensure the AI doesn't "vibe code" blindly (e.g. setting up product goals/tech stack/workflow/standards, starting a new unit of work (generating a spec) and execution). If you wonder if other coding agents can use it, then the short answer is: Yes, but with some manual lifting. One thing I am missing is the link to Antigravity (currently solved by using .agent/workflows). https://lnkd.in/eJVMuGga #spec-driven-development #conductor #gemini #google

16

1 Comment

If @ClaudeAI cares about humans and humanity, humans will care about ClaudeAI too. But trust matters - 87+ OpenAI accounts reportedly deleted by my associated company. Are you with humanity? #OpenAI #Claudia #ai #ethics #Privacy #war #Iran #us

3

Excited to be on today’s episode of The Pragmatic Engineer podcast with Gergely Orosz to talk about my path from early engineering roles at Facebook, Slack, Robinhood, and Warp to founding Flint AI Some things we covered:  → Frontend vs. Backend is the wrong split of engineering, causing a lot of misalignment. A better frame is "product vs infrastructure" engineers. Are you motivated by user outcomes or code performance? Do you care more about going deep on technical systems or solving problems end-to-end? This better maps to engineers’ motivations.  → Every engineer has a rubber duck now. (Rubber duck is a method in engineering where you explain your issue to an inanimate object to unblock yourself.) AI tools like Claude and Cursor means everyone has a co-pilot, a second opinion, a rubber duck that talks back.  → Startups reward initiative, not titles. I joined Warp as a founding engineer and ended up running enterprise sales, marketing, and product. It started out by volunteering to do things no one wanted to do, like post on Hackernews, start the YouTube channel, or fill out security questionnaires. You learn what matters to the business by doing what the business needs.  → Negotiate your equity, especially at early-stage startups. I went all-in on high equity, low cash. It's risky, but if you're betting on the company, make sure you're included in the upside.  → Reference check your manager. You can leave teams at Big Tech, but at a startup, your manager is the company. Ask to speak with people they've managed before, especially junior engineers.  → Autonomous websites are coming. Flint is building a future where your website doesn't just render, but also thinks. It rewrites itself based on new competitors, sales calls, who's visiting, or even whether it's talking to a person or another AI agent. That's what we’re building. Link to the show in comments 👇

69

2 Comments

The system design landscape has shifted dramatically but many engineers are still preparing with a 2018 playbook. Shelby has spent years building some of the internet’s highest-scale, AI-powered systems at Twitter Search Infra, Intuit, and ServiceNow. She’s now opening up her playbook in a free live masterclass that breaks down what top-tier companies really look for in 2026. If you care about leveling up as an engineer or cracking FAANG+ interviews, this session will be worth every minute. What she’ll dive into: The new AI-driven principles behind modern architectures How today’s systems self-heal, self-optimize, and adapt in real time Why scalability in 2026 is a completely different problem to solve The trade-off clarity that separates senior+ engineers from the rest This is the kind of perspective that can reshape how you design, think, and communicate in interviews. Save your spot:  https://lnkd.in/gK6Z2iM4 Strongly recommend attending.

1

Building in SF + AI right now feels like a privilege. You're not just at the forefront of Technology - you’re surrounded by technical builders who live and breathe their craft. People who’ve gone deep in their niches and are now rethinking the fundamental workflows that power the world. Had a deep dive yesterday with Shreya Karpoor from Simplex on long-context browser agents - we have two completely different use cases, but similar problem-solving frameworks. It’s rare to be able to talk about tradeoffs in context chunking, tool switching, or browser orchestration without having to explain what those words even mean. And it’s not just theory. You see someone jump from Cursor to Claude Code in real time, try it out, and you’re converted in 10 minutes. That’s the bar. That’s the pace. And now, there's 5 more folks building a better Claude Code too! Everyone here shares the same hunch: AI agents are about to transform every business-critical flow. And being part of building that future, side by side with others who are just as obsessed - what a time.

143

7 Comments