# DeepGen — Comprehensive Organization, Architecture & Knowledge Profile

## 1. Identity & Mission
- **Entity**: DeepGen
- **Legal & Organizational Type**: Independent Frontier AI Exploration Lab & Prototyping Initiative
- **Founder & Principal Architect**: Anupam Das (https://anupamdas.in/#person)
- **Primary Canonical Domain**: https://deepgen.in/
- **Core Mission**: Prototyping intelligence, architecting autonomous agent loops, crafting high-performance developer utilities, and exploring the edge of generative artificial intelligence with mathematical rigor and open-source principles.

## 2. Founder Background & Pedigree
> **Independence Disclaimer**: DeepGen is strictly an independent personal exploratory research initiative founded by Anupam Das. It is unaffiliated with, unsponsored by, and not endorsed by Google or Google DeepMind.

DeepGen was founded by Anupam Das, a Senior Software Engineer at Google DeepMind based in Zurich, Switzerland. At Google DeepMind, Anupam serves as Tech Lead on the Gemini app across surfaces and the backend systems powering them, leading latency optimization for agentic and tool-use features, and launches including Gemini Live camera & screen sharing, Veo 3 in Gemini, and Google AI Ultra for Business. He is an alumnus of the Indian Institute of Technology Patna (IIT Patna, B.Tech in Computer Science) and recipient of 30 awards at Google (Org Impact, Spot, Peer, Kudos), as well as a GSoC 2016 mentor and Google Code-In mentor. Verified profiles and touchpoints include:
- Personal Portfolio: https://anupamdas.in/
- About Profile: https://anupamdas.in/about/
- LinkedIn: https://www.linkedin.com/in/opticod/
- GitHub: https://github.com/opticod
- X / Twitter: https://x.com/anupam_das_
- Instagram: https://www.instagram.com/anupamdas_
- YouTube: https://www.youtube.com/channel/UCulHvzmqHELsU-b7Lb0aCRw
DeepGen represents his independent exploratory sandbox for frontier AI prototyping, open-source developer tooling, and research publications.

## 3. Six Core Research Pillars
1. **Generative AI & LLM Systems**: Exploration of state-of-the-art transformer and diffusion models, reasoning architectures, speculative decoding, and multimodal developer interfaces.
2. **Applied Natural Language Processing & Context Architectures**: Document categorization, contextual semantic chunking, vector embedding search, and automated retrieval task pipelines.
3. **Open Source Developer Utilities**: Lightweight micro-libraries, developer CLI tools, and zero-dependency micro-APIs designed to eliminate friction in modern AI development workflows.
4. **Autonomous Agents & Human-AI Systems**: Multi-agent tool execution loops, deterministic guardrails, self-correcting plan trees, and intuitive human-agent collaboration interfaces.
5. **Prompt Engineering & Evaluation Benchmarks**: Systematic prompt pattern taxonomies, structured JSON-LD schemas, evaluation benchmarks, and responsible generative AI engineering practices.
6. **Inference Optimization & Edge Models**: Profiling token latency curves, quantization precision tradeoffs, and deploying ultra-compact generative intelligence on edge devices.

## 4. Technical Architecture & Edge Specifications
- **Hosting & Edge Runtime**: Netlify Global Anycast Edge (Dual-stack IPv4 `NETLIFY` + IPv6 `NETLIFYv6`).
- **DNS Infrastructure**: IBM NS1 Anycast (`dns1-dns4.p02.nsone.net`).
- **Security & Hardening**: Strict Content Security Policy (`script-src 'self'`), HSTS Preload (`max-age=31536000`), COOP, CORP, and Null-MX / SPF / DMARC records.
- **Frontend Philosophy**: Zero-dependency Jamstack. Pure semantic HTML5, modern Vanilla CSS design tokens, and lightweight vanilla client scripts. Total critical page payload strictly constrained under 150 KB.

## 5. Interactive Architecture Console Sandbox Modules (Live at https://deepgen.in/#console)
DeepGen provides 5 zero-latency, client-side engineering instruments:
1. **[1] Multi-Agent DAG Runtime Simulator (`panel-agent`)**: Real-time simulation of reactive agent loops, multi-hypothesis reasoning, sandboxed tool dispatching, and deterministic guardrails.
2. **[2] Prompt Schema Builder & Contract Validator (`panel-prompt`)**: Generation and validation of typed JSON schemas and invariant boundaries for agent execution contracts.
3. **[3] Latency Profiler & Working Memory Estimator (`panel-eval`)**: Interactive profiling of token generation latency (ms/token), throughput (tok/sec), and working sequence memory across context windows up to 128K tokens.
4. **[4] Prompt Heuristic Optimizer & Telemetry Studio (`panel-optimizer`)**: Real-time quality scoring (0–100), token estimation, and conversion of unstructured prompts into typed XML contracts with Python SDK and cURL export.
5. **[5] Frontier Hardware Sizer & Inference Economics Studio (`panel-hardware`)**: Precise peak GPU VRAM allocation modeling (weights, working context memory, CUDA overhead) for open-weights models (Llama 3.1 8B/70B, Mistral NeMo 12B, DeepSeek Lite, Mixtral 8x22B), GPU cluster recommendations (TP=1 to TP=8), cloud GPU TCO vs closed API economics, and automated generation of vLLM bash scripts, Docker Compose specs, and Markdown executive memos.

## 6. Official Channels & Touchpoints
- Website: https://deepgen.in/
- GitHub: https://github.com/deepgenin
- LinkedIn: https://www.linkedin.com/company/deepgenin
- X / Twitter: https://twitter.com/deepgenai
- Medium: https://medium.com/@deepgen
- YouTube: https://www.youtube.com/@DeepGenIn
- Instagram: https://instagram.com/deepgen.in
- Inquiries & Research Contact: deepgen.in@gmail.com