const tanmay = {
name: "Tanmay Raut",
role: "software developer @ VisioneerIT",
location: { base: "Pune, IN", hours: "EST", team: "Baltimore, MD" },
building: ["AltamIQ", "Industrial Briefs", "OryonIQ QA"],
stack: {
web: ["Next.js", "React Native", "Express"],
data: ["PostgreSQL", "Docker", "n8n"],
},
previously: "founding engineer @ BrainChain",
openTo: ["full-stack", "AI engineering"] as const,
caffeineLevel: "depends",
stillLearning: (): true => true,
} satisfies Developer;sole engineer at visioneerit, building from pune on est hours. ship things end-to-end, own the stack, keep the bug tracker honest. started out as founding engineer at brainchain, learned a lot getting thoughtpro onto the play store, and have been chasing that same "ship it, watch it break, fix it, ship it again" loop ever since. still mostly self-taught, still figuring it out, still building.
- altamiq — partnership-management platform on next.js, react, express, postgres, docker, digitalocean. three-tier rbac, licensing server, the whole thing.
- oryoniq — pre-production qa on a contractor platform. breaking it before customers do.
- industrial briefs — 9-agent autonomous newsroom on n8n, supabase, and ghost. runs itself, mostly.
- internal r&d — turning ai prototypes into production-ready deployments. less demo, more uptime.
Industrial Briefs @ VisioneerITfully autonomous AI newsroom. 9-agent pipeline that scrapes, fact-checks, writes, and publishes ~15 articles a day across AECM sectors. zero humans in the loop after deploy. tech: n8n · Supabase · Ghost CMS · LLM orchestration
|
AltamIQ @ VisioneerITpartnership-management platform with three-tier RBAC and a Docker-based licensing server so the whole thing can be self-hosted per client. multi-tenant from day one. tech: Next.js · React · Express · Postgres · Docker · DigitalOcean
|
ThoughtPro @ BrainChainhealthcare app shipped to the Play Store. 20+ screens, 95% iOS/Android code reuse, full i18n across Hindi, Marathi, English. tech: React Native · TypeScript · SQLite · JWT |
production-grade virtualized data grid. holds 60fps at 1000+ rows using TanStack Virtual, with inline AI enrichment running on Groq + Llama for sub-second cell completions. tech: Next.js · TanStack Virtual · Groq · Llama 3
|
|
self-hosted AI workspace. local-first, no vendor lock-in, runs your own models and your own context. tech: Next.js · local LLM runtimes · vector storage |
restaurant site with a 120-frame scroll-driven cinematic intro, bilingual (English/Marathi), and a 100/100 Lighthouse across the board. tech: Next.js · scroll-triggered canvas · i18n
|
languages
frontend
backend & APIs
databases & storage
AI & automation
DevOps & cloud
tools
odysseus is a self-hosted ai workspace (21k★). been chipping away at the security and self-hosting edges. 15 reported issues closed across 9 PRs.
merged into upstream:
- #782 — fix TOCTOU race in chat stream status endpoint. closes #658
- #776 — fix searxng container permission errors during setup. closes #721
rebased + answered, waiting on maintainer:
- #784 — exempt task webhook trigger from session auth. closes #621 (+ bonus: #640, #643)
- #785 — honor
AUTH_ENABLED=falsein route-level auth gate. closes #622 (+ bonus: #634, #804) - #786 — stop wiping session model on endpoint delete, support in-place updates. closes #587
- #783 — lift deep-research hard timeout into a setting. closes #595 (+ bonus: #731)
- #769 — support extra CA bundle for private-CA LLM providers. closes #722
- #763 — apply SafeSearch by default across search providers. closes #718
queued for first review:
- #759 — expose
manage_notesvia native function calling. closes #174, #713
full contributor history on odysseus
open to remote full-time roles with EST overlap. immediate joiner, startup-friendly, end-to-end ownership is the default. fastest reply over email or linkedin DMs.
built in public. shipped on purpose.