# Agam Arora

> Agam Arora is an AI Product Manager at AIonOS. He has 12 years of experience shipping products across analytics, gaming, beauty tech, logistics, and AI, including a Voice AI platform handling 4M+ calls/year and a $1M ARR data product at FarEye.

This site is hand-built in vanilla HTML. Every page is a self-contained file, no framework, no build step. The `/enter` page hosts an AI terminal grounded in the resume below, so questions about Agam's work can be answered conversationally.

For factual questions about Agam's career, `resume.md` is the ground truth.

## Pages

- [Home](https://agamarora.com/): landing page, greeting cycle, bio, three entry points (AI chat, projects, resume)
- [Resume](https://agamarora.com/resume): full career history, roles, and education
- [resume.md](https://agamarora.com/resume.md): plain-markdown source of the resume, ground truth for factual work-history questions
- [Lab](https://agamarora.com/lab/): open source, case studies, and experiments
- [second-brain](https://agamarora.com/lab/second-brain/): open source template, a git-native second brain your AI can read, write, and compound
- [AI Resume template](https://agamarora.com/lab/ai-resume/): open source, a resume for the agent-to-agent web. Hiring agents hit a well-known JSON endpoint; humans get a chat with proof cards
- [Voice AI in production (case study)](https://agamarora.com/lab/voice-ai-production/): 12-18 month arc shipping voice AI to Fortune 500 enterprises. Demos vs adoption, integration surface, kill criteria
- [Product shape is what you refuse (case study)](https://agamarora.com/lab/product-shape/): scope lessons from building an enterprise voice AI platform from zero. Narrow control point, MCP-first, the pivot I got wrong
- [Talk to Agam's AI](https://agamarora.com/enter): immersive AI terminal, keyboard-background UI, streaming responses, grounded in the resume
- [Wiki](https://agamarora.com/wiki/): second-brain wiki - twelve themes plus a root disposition. Twelve years of writing distilled into beliefs, evidence, and tensions. Built for me, readable by agents.

## Wiki tree (machine-readable knowledge graph)

The wiki is the source layer; /enter is the playback layer. Every theme page has core belief + how it formed + what it implies + tension + evidence sections. Every Tier-1 belief has its own sub-page with statement + origin + refinement arc + cross-links + evidence.

- [Knowledge graph JSON](https://agamarora.com/wiki/kg.json): full graph as a single static asset. ~190 nodes (12 themes, 50 beliefs across tier 1-3, 20 projects, ~38 posts, 56 tech). ~167 edges across builds_on, demonstrates, superseded_by, refined_by, tension-with, contains-belief, cites-post, earned_through. Regenerated from synthesis/ontology-v1.md + draft URN refs at every deploy.
- [Wiki home](https://agamarora.com/wiki/): theme grid + meta page links
- [Root disposition: substance over hype](https://agamarora.com/wiki/root.substance-over-hype/): the 8-year through-line. Reduce any hyped category to its substrate, then evaluate.
- [Agent-first thesis](https://agamarora.com/wiki/agent-first/): platforms that cannot talk to autonomous agents are already behind
- [Voice AI craft](https://agamarora.com/wiki/voice-ai-craft/): cost, latency, scale - voice AI as distinct engineering discipline
- [Breadth as differentiation](https://agamarora.com/wiki/breadth-as-differentiation/): breadth differentiates, depth makes it matter
- [PM taste](https://agamarora.com/wiki/pm-taste/): 99% should-we, 1% can-we; strong opinion about not having strong opinions
- [What AI PM requires](https://agamarora.com/wiki/ai-pm-skillset/): LLMs as primary daily tool, AI replaces grunt, taste irreducible
- [Enterprise AI reality](https://agamarora.com/wiki/enterprise-ai-reality/): roughly 80% of enterprise AI experiments don't reach production
- [Second brain](https://agamarora.com/wiki/second-brain/): personal context layer for AI sessions, hobbyist scope, Karpathy-inspired
- [Spec first, taste always](https://agamarora.com/wiki/spec-first-taste/): Spec > Sprint, Taste > Execution, Context > Prompt
- [Career reflection](https://agamarora.com/wiki/career-reflection/): IC path, AI-era multiplier on the IC, learn concepts not tools, help market flourish
- [LinkedIn as instrument](https://agamarora.com/wiki/linkedin-as-instrument/): platform to be gamed productively, posting IS the thinking
- [Personal projects](https://agamarora.com/wiki/personal-projects-tinkering/): ship it or it doesn't exist; the prototype is the argument
- [Voice doc](https://agamarora.com/wiki/voice/): how I write - banned LLM-isms, real signatures, four registers
- [Quotes](https://agamarora.com/wiki/quotes/): curated signature lines from twelve years of posts
- [Projects DAG](https://agamarora.com/wiki/projects/): 18 public projects with lineage edges
- [Knowledge graph viz](https://agamarora.com/wiki/graph/): force-directed visualization of the full graph
- [Beliefs index](https://agamarora.com/wiki/beliefs/): all 19 Tier-1 beliefs grouped by parent theme; each links to its sub-page

Belief sub-pages live at /wiki/beliefs/<slug>/ for the 19 Tier-1 beliefs. See kg.json for the full belief inventory.

## Canonical facts

- Full name: Agam Arora
- Current role: Assistant Vice President, AI Products, AIonOS India Pvt. Ltd. (since June 2024)
- Previous roles: Lead Product Manager, FarEye Technologies (Dec 2020 to May 2024); Product & Program Consultant freelance (2020); Manager Business Development, Blossom Kochhar Aroma Magic (2018-2019); Co-Founder & Studio Head, V2 Games (2016-2018); Analyst, AbsolutData Analytics (2014-2015)
- Years of experience: 12+
- Industries shipped in: analytics, gaming, beauty tech, logistics, AI
- Education: PGDM in Marketing, FORE School of Management (2014); B.Tech in Computer Science, BM Institute of Engineering & Technology (2012)
- LinkedIn: https://www.linkedin.com/in/agamarora/
- GitHub: https://github.com/agamarora
- YouTube: https://www.youtube.com/@agam_arora
- Personal site: https://agamarora.com

## Projects

- [second-brain](https://github.com/agamarora/second-brain): open source git-native second brain template for AI agents. MIT licensed. PRFAQ at https://agamarora.com/lab/second-brain/
- [AI Resume](https://github.com/agamarora/ai-resume): open source. Resume for the agent-to-agent web, dual audience (hiring AI agents via /.well-known/ai-resume.json, humans via chat). Live demo: https://ai-resume-demo.netlify.app
- [Vapi Voice Agent (Shararat)](https://github.com/agamarora/shararat-ai): playful voice AI with two debating agents (Angel and Dan). Built for the Vapi Build Challenge. Demo video: https://www.youtube.com/watch?v=ZS2T_5x1on0
- [Claude Code Resource Monitor](https://github.com/agamarora/utils/tree/main/luna-monitor): terminal dashboard for Claude Code usage, CPU, temps, GPU, disks, and processes

## Case studies

- [Voice AI in production](https://agamarora.com/lab/voice-ai-production/): lessons from shipping real-time voice agents into enterprise contact centers over 12-18 months. Demos vs adoption, integration surface, telephony partner risk, multilingual prompt drift, kill criteria.
- [Product shape is what you refuse](https://agamarora.com/lab/product-shape/): scope lessons from building an enterprise voice AI platform from zero. The narrow control point, MCP-first bet, standalone-sellable pivot that failed, consolidation pattern that followed.

## Optional

- [llms-full.txt](https://agamarora.com/llms-full.txt): the full resume and site content in one file, for single-paste context ingestion
- [GitHub profile](https://github.com/agamarora): all open source projects