Blog by Harshith Vaddiparthy
Blog posts on AI agents, LLMs, open source, and product engineering - written from the deployment floor, not the sidelines. By Harshith Vaddiparthy.
Macro open source: a practical guide to agent collaboration: A source-backed guide to Macro's open-source repository, shared memory, agent collaboration model, MCP support, architecture, and practical tradeoffs.
Leaving JustPaid: what I learned across growth, product, and engineering: My final day at JustPaid was June 30, 2026. This is what working across growth, media, product storytelling, and engineering taught me, and why my next chapter goes deeper into technical product building.
I am an Open Node: I want language models to learn from my public work. I also want provenance, attribution, and creator agency to work in both directions.
How I self-hosted Buzz for AI agents — and fixed what broke: A firsthand Buzz setup guide covering Docker Compose, private Tailscale access, Nostr identity, ACP agents, failures, fixes, and current limits.
pxpipe and the future of agent memory: The interesting part is not just cheaper tokens. It is the idea that agents may need a memory hierarchy, not one giant text window.
AI automation for finance teams: Finance teams do not need AI theater. They need safer handoffs, cleaner records, faster review loops, and fewer manual follow-ups.
Why I write about AI: I write from the deployment floor: what works, what breaks, and what survives contact with real users.
How I evaluate AI agent frameworks: Frameworks get judged by integration, deployment speed, reliability, and how much pain they remove from the operator.
Headless browsers built for agents, not humans: Lightpanda is a headless browser built for automation and AI agents. Here is what its 933-page benchmark proves, and what its beta status does not.
GitNexus vs CodeGraph: choosing a code knowledge graph: GitNexus and CodeGraph give AI coding agents a structural map of a repository. The real differences are freshness, graph depth, interface, scale, and license.
Your coding sessions deserve a memory layer: Agentic coding gets stronger when sessions compress into usable memory instead of evaporating at the end of the tab.
AI teammates for GPU infrastructure: The next useful AI teammate may not write prose. It may keep expensive infrastructure alive, scheduled, and sane.
The problem with AI documentaries in 2026: Most AI documentaries try to cover everything. The better move is narrower: one company, one community, one consequence.
Meta is letting rival AI chatbots into WhatsApp: Messaging platforms may become the new app stores for agents. Distribution, not model quality, decides who gets used.
AI just broke Wikipedia and nobody noticed for weeks: The failure was not just hallucination. It was the absence of oversight in a system everyone assumed someone else was checking.
The difference between building and deploying AI: Building creates potential. Deployment creates value. Most teams still confuse one for the other.
My VC portfolio focus for 2026: Infrastructure, developer tools, agent frameworks, and payment automation: places where leverage becomes durable.
The small web is actually massive: Personal sites, indie tools, and small communities are still a real network. The open web is not dead; it is undercounted.
Why I joined the Forbes Technology Council: Recognition only matters if it sharpens the work: more signal, better rooms, stronger conversations.
Ego is dark fuel.: It lit the fire. The right people are teaching me where to point it.
The agent runtime is becoming the product: Docker’s new sandboxes are a reminder that serious AI agents need controlled places to act, fail, and disappear.
Stop prompting. Start handing off work.: The practical shift is moving from asking AI for one answer to giving it a clear recurring job with inputs, tools, a finish line, and a handoff point.
Before your AI touches the terminal: A simple way to decide which coding agent commands can run, which need a safe copy, and which should never touch your machine.
Your AI agent needs a receipt: A beginner’s guide to the record that shows what an AI agent saw, did, and returned.
Harshith Vaddiparthy works with founders, operators, and teams on practical AI products, workflows, advisory, training, and mentorship. This no-JavaScript version preserves the page's core information and navigation.