An always-on AI desktop overlay — live in production, with paying customers — that captures screen and audio context to debug code, assist in live meetings, and dictate Wispr-style, triggered from anywhere by a shortcut to cut tool-switching. It runs a local Whisper-style STT pipeline with ~500ms–1.5s partial-transcription latency and a RAG layer over pgvector/Qdrant. Reasoning is pluggable across OpenAI, Anthropic, Gemini, or local models with bring-your-own-key support. Privacy-first: avoids storing raw screenshots/audio by default, with an in-app permissions view.
"}},{"node":{"frontmatter":{"title":"SERO","cover":{"childImageSharp":{"gatsbyImageData":{"layout":"constrained","placeholder":{"fallback":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAALCAYAAAB/Ca1DAAAACXBIWXMAAAsTAAALEwEAmpwYAAAA9ElEQVQoz62SbUsDMRCE8///pCgFP4h65pLdbPZlJLm2XhWsLR4M8xA2k0320sPjAWKBJVe8LXmKRSHqaN2m/9BxvYmidYV5TFUWJGqCp8MzXl6X84YR9CeN2n29KNJaGblU5LWCWKDmU10dfc96nUczyeYmA9eCnDMiAuMbfiubO5J7QKRjzQuY+aLwFiG2dzx3WEv+n0CPABHDTJE/3lHp/tCtQw80Ebg7pBGI6BgIxN5/5e0dj4EOVd1OcUUtZYbfo20oARCPK9s8pVFB4zqnTmWdTHW9wgQqGaqGNP6hceVToDRGlza9XTB9MX9nmT6G+wlDqWIFDd5OygAAAABJRU5ErkJggg=="},"images":{"fallback":{"src":"/static/ba4fcb612bbd72427ace8cd871792c0c/bc3b9/sero.png","srcSet":"/static/ba4fcb612bbd72427ace8cd871792c0c/9a130/sero.png 175w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/a3fa1/sero.png 350w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/bc3b9/sero.png 700w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/87926/sero.png 1400w","sizes":"(min-width: 700px) 700px, 100vw"},"sources":[{"srcSet":"/static/ba4fcb612bbd72427ace8cd871792c0c/dae43/sero.avif 175w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/d7667/sero.avif 350w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/7ec1a/sero.avif 700w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/71a30/sero.avif 1400w","type":"image/avif","sizes":"(min-width: 700px) 700px, 100vw"},{"srcSet":"/static/ba4fcb612bbd72427ace8cd871792c0c/5d873/sero.webp 175w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/26a00/sero.webp 350w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/f23f0/sero.webp 700w,\n/static/ba4fcb612bbd72427ace8cd871792c0c/2c2d0/sero.webp 1400w","type":"image/webp","sizes":"(min-width: 700px) 700px, 100vw"}]},"width":700,"height":394}}},"tech":["Go","GraphQL","PostgreSQL","React","BetterAuth"],"github":"","external":"https://app.squiwo.com/","cta":"https://app.squiwo.com/"},"html":"Multi-tenant manufacturing-scheduling SaaS (live in production) with a TypeScript/React frontend, Go/GraphQL backend, and PostgreSQL — organization-level RBAC via BetterAuth, onboarding, orders, and subscription billing through Dodo Payments. Its core is a heuristic DAG-based production scheduler (topological ready-queue with in-degree release, critical-path ranking, capacity-aware interval slotting, greedy machine selection, idle-gap backfill) that generates full production plans in minutes instead of hours.
"}},{"node":{"frontmatter":{"title":"ATS Beater","cover":null,"tech":["FastAPI","Vue 3","PostgreSQL","Google Gemini","LaTeX"],"github":"","external":"https://resume.debugblackbox.com/","cta":"https://resume.debugblackbox.com/"},"html":"Open-source AI resume-tailoring SaaS (173 unit tests, deployed on Google Cloud Run) built with FastAPI, Vue 3, PostgreSQL, and LaTeX/pdflatex. A two-phase AI pipeline on Google Gemini (Flash for profile structuring, Pro for keyword-tailored generation) drives generation, while Google ADK chat agents perform JSON-Patch edits and background PDF recompilation. Includes a Razorpay-backed credit system (daily-free, packs, time passes) with synchronous deduction and automatic refunds on failure, multi-tenant org labeling, and an admin panel.
"}}]}}}