Volunteer Experience
-
Volunteer
g0v
- less than a year
g0v builds public-interest technology by “forking” government - taking official datasets, services, and information resources and rebuilding them into more accessible, useful tools for citizens.
-
Co-Director
LexChat
- 1 year
LexChat raised $120K+ for social impact orgs and built forums for ongoing civic dialogue.
Other similar profiles
Explore more posts
-
Vineet Chirania
Tech stacks can be learned. System thinking, ownership & hunger cannot. That is what makes hiring engineers so hard. 100% agreed with Brian (Coinbase CEO) here, most of the times best hires are totally unqualified on paper. After building two startups and hiring across both, the pattern is clear. The best people are problem solvers who want to do something big and genuinely believe in the mission. They take initiative without being asked, stay curious, and never stop learning. But the one thing that matters more than all of this is trust. Trust cannot be tested in an interview. It is earned slowly over time through consistent actions, not words. The best hires are the ones who build that trust in the first few months by simply doing what they said they would do. This is exactly how we are building the team at CubeAPM. If you are hiring engineers, I would love to hear what filter works best for you.
1 Comment -
Jeffrey Morelli
One of our top code contributors last month was not an engineer. It was Rosey Li from our Product Ops team. Most companies are using AI to help engineers write code faster. We took a different approach. We leverage AI to make more people engineers. A few weeks ago, two of our engineers spent a weekend building Silna Forge, a system of remote coding agents that runs on a fully self-hosted version of our application. The goal wasn’t just speed, it was creating leverage across the organization. The first real test came quickly. We needed to implement a new billing configuration to account for a new pricing structure, the kind of request that would normally sit in a backlog for weeks. Instead of entering the product roadmap, our Chief of Staff generated the feature in a secure sandbox, refined it with the agent, and had it reviewed and merged in under an hour. What used to take weeks moved from idea to production in a single sitting. Since launch, Forge has generated 57% of our merged pull requests, primarily across internal tooling and operational enhancements. The people closest to day-to-day friction can now build the tools they need, rather than routing every improvement through a queue. The shift is both technical and organizational. Internal tools rarely have hard ROI metrics. There’s no churn signal tied to a missing dashboard filter. No revenue tracking on a billing configuration tab. So when it comes time to allocate engineering capacity, internal tooling almost always loses to revenue-generating features, even when the operational drag it would resolve is real. Most inefficiency in companies lives in the spaces between teams, in handoffs, approvals, and backlog queues. Engineers should be pushing boulders, not moving pebbles. Silna Forge handles the pebbles, reducing context switching and clearing backlog noise for our engineering team. When you reduce the number of touchpoints between identifying a problem and solving it, you move faster but also function cleaner. We operate in a regulated healthcare environment, where plugging in third-party AI tools isn’t an easy option. Instead, these constraints forced us to design something in-house on top of our own infrastructure that works even better than what we see in the market. Looking forward to seeing what my team builds next with Silna Forge.
3 Comments -
Gobhanu Sasankar Korisepati
6 weeks into YC, here's something I didn't expect: The hardest part of building an AI product isn't the AI. It's convincing people to trust it with something that feels personal. Scheduling sounds simple. But every meeting has stakes. A botched interview loop can lose a candidate. A double-booked exec call can kill a deal. People don't want "AI scheduling." They want to never think about scheduling again, AND never get burned. That's a trust problem, not a tech problem. So we stopped selling features and started selling reliability. Every Vela interaction is designed to feel like a great EA handled it!
8 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content