# Preferred Networks (PFN · プリファードネットワークス) — Complete Application Dossier & LinkedIn Kit

- **Applicant:** Péter József Lódri (`LODRI PETER JOZSEF` · ロドリ・ペーテル・ヨージェフ)
- **Target Role:** Senior Rust Systems & Custom AI Accelerator Engineer (MN-Core / PLaMo / Rustuna)
- **Location:** Otemachi, Chiyoda-ku, Tokyo (大手町フィナンシャルシティ)
- **Compensation Band:** `¥12,000,000 – ¥22,000,000+ JPY`
- **Dedicated Live Subdomain:** [https://pfn.vaked.dev](https://pfn.vaked.dev)
- **Formal Résumé PDF:** [https://pfn.vaked.dev/resume.pdf](https://pfn.vaked.dev/resume.pdf)
- **Engineering Portfolio PDF:** [https://pfn.vaked.dev/portfolio.pdf](https://pfn.vaked.dev/portfolio.pdf)
- **LinkedIn Profile:** [https://www.linkedin.com/in/peter-lodri/](https://www.linkedin.com/in/peter-lodri/)
- **GitHub & OSS Architect:** [github.com/peterlodri-sec](https://github.com/peterlodri-sec) · [github.com/8b-is](https://github.com/8b-is) · [osaurus-ai/osaurus](https://github.com/osaurus-ai/osaurus) · [simplifaisoul/osiris](https://github.com/simplifaisoul/osiris) · [headroomlabs-ai/headroom#1418](https://github.com/headroomlabs-ai/headroom/pull/1418)

---

## Direct Application & LinkedIn Contact Directory

- **Official Careers Portal:** [https://www.preferred.jp/en/careers/](https://www.preferred.jp/en/careers/)
- **Direct Recruiting Email:** `careers@preferred.jp`
- **LinkedIn Company Page:** [https://www.linkedin.com/company/preferred-networks/](https://www.linkedin.com/company/preferred-networks/)
- **LinkedIn People Directory:** [https://www.linkedin.com/company/preferred-networks/people/](https://www.linkedin.com/company/preferred-networks/people/)
- **LinkedIn Recruiter & Team Search:** [Find Preferred Networks (PFN · プリファードネットワークス) Recruiters & Engineers](https://www.linkedin.com/search/results/people/?keywords=Preferred%20Networks%20%28Recruiter%20OR%20Talent%20OR%20HR%20OR%20MN-Core%20OR%20PLaMo%29)

### Key Leadership & Talent Contacts
- **Toru Nishikawa (西川徹)** (Co-Founder, President & CEO, Preferred Networks): [https://www.linkedin.com/search/results/people/?keywords=Toru%20Nishikawa%20Preferred%20Networks](https://www.linkedin.com/search/results/people/?keywords=Toru%20Nishikawa%20Preferred%20Networks) — Co-founder leading PFN's vertically integrated computing stack (MN-Core, PLaMo, Matlantis).
- **Daisuke Okanohara (岡野原大輔)** (Co-Founder & Chief Technology Officer (CTO), Preferred Networks): [https://www.linkedin.com/search/results/people/?keywords=Daisuke%20Okanohara%20Preferred%20Networks](https://www.linkedin.com/search/results/people/?keywords=Daisuke%20Okanohara%20Preferred%20Networks) — Co-founder & CTO pioneering deep learning infrastructure, low-bit inference, and PLaMo LLMs.
- **PFN Talent Acquisition & Compiler/Systems Team** (Engineering Recruiting · Otemachi, Tokyo): [https://www.linkedin.com/search/results/people/?keywords=Preferred%20Networks%20Recruiter%20OR%20Talent%20Tokyo](https://www.linkedin.com/search/results/people/?keywords=Preferred%20Networks%20Recruiter%20OR%20Talent%20Tokyo) — Technical recruiting team for MN-Core runtime, Optuna/Rustuna, and PLaMo inference.

### LinkedIn Connection Note (< 300 chars)
```text
Hi! I built a live interactive MN-Core systolic array & Rust systems portfolio specifically for Preferred Networks in Otemachi: https://pfn.vaked.dev — showcasing bit-exact C/Rust 1.58-bit ternary codecs (pureQTern) & custom GPU kernels. Would love to connect! — Péter
```

### LinkedIn InMail / Direct Email Body
```text
Dear Preferred Networks (PFN) Engineering & Talent Team,

I created a dedicated interactive systems portfolio and bilingual application dossier for Preferred Networks in Otemachi, Tokyo:

• Live Interactive Portfolio (MN-Core Systolic Array Simulation): https://pfn.vaked.dev
• Formal 履歴書 + 職務経歴書 (PDF): https://pfn.vaked.dev/resume.pdf
• Systems Portfolio (PDF): https://pfn.vaked.dev/portfolio.pdf
• Full Technical Application Note: https://pfn.vaked.dev/application

Why PFN is my ideal technical home in Tokyo:
• Bit-Exact C/Rust Low-Bit Runtimes: I built pureQTern and project-zero — dependency-free 1.58-bit ternary codecs in C and Rust verified to the bit by golden SHA-256 hashes, plus custom Metal GPU kernels (MLX-QUANT) achieving 12.80× compression.
• High-Throughput Systems & Open Source: Core Architect on Osaurus (osaurus-ai/osaurus) & Osiris (simplifaisoul/osiris), contributor to Headroom (PR #1418), and 9+ years leading high-concurrency Go/Rust/Python systems (ASAPP, DailyWire).

Warm regards,
Péter József Lódri
https://www.linkedin.com/in/peter-lodri/
```

---

# 03 — Preferred Networks / PFN (Tokyo, Otemachi)
**Role:** Senior / Principal Rust Systems & Custom AI Accelerator Engineer (`MN-Core` / `Rustuna` / LLM Inference Runtime)  
**Estimated Compensation:** ¥12,000,000 – ¥22,000,000+ JPY (Tokyo Hybrid / On-Site)  
**Application Portal:** [https://www.preferred.jp/en/careers/](https://www.preferred.jp/en/careers/)  
**Resume to Attach:** [`../teamlab/lodri-resume.pdf`](../teamlab/lodri-resume.pdf) & [`../resumes/cv-peter-lodri-taiko.pdf`](../resumes/cv-peter-lodri-taiko.pdf)

---

## 1. ELI5: What Does Preferred Networks (PFN) Do & Why Do They Need You?
- **What they do:** PFN is Japan's premier deep-tech hardware + AI company. They design their own supercomputer AI chips (**MN-Core / MN-Core 2**), build foundation models (**PLaMo**), run materials-discovery simulations (**Matlantis**), and rewrite slow Python AI tooling into ultra-fast **Rust** (such as **Rustuna**, their Rust engine for Optuna).
- **Why you are a SOTA match:** PFN's core engineering bottleneck is squeezing maximum compute density out of custom silicon and memory hierarchies using **Rust, C, SIMD, and low-bit quantization**. Your `BFP-64` bitplane packing (`popcnt(pos & inp) - popcnt(neg & inp)`), `wupz` 5-trits-per-byte codecs, and zero-dependency C/Rust ternary inference engines speak directly to PFN's compiler and runtime engineers.

---

## 2. Ready-to-Paste Application Cover Letter / Statement of Purpose

```text
Dear Preferred Networks Engineering & Recruiting Team,

I am applying for a Senior Systems / AI Infrastructure Engineer role at Preferred Networks in Tokyo. Over the past 8+ years, I have built high-throughput distributed backends (ASAPP, Virality Data, DailyWire, Plandek) and low-level Rust/C/Swift compute engines focused on low-bit quantization, deterministic numeric kernels, and hardware-accelerated inference.

Why my work aligns directly with PFN's MN-Core, PLaMo, and Rust systems stack:

1. Bitplane Quantization & Hardware-Sympathetic Kernels (`8b-is/cogito`, `pureQTern`):
   - Designed and shipped dependency-free C and `#![no_std]` Rust inference runtimes for 1.58-bit ternary (`{-1, 0, +1}`) models, replacing floating-point MAC arrays with two-plane 64-trit POPCNT dot products (`BFP-64`) and 5-trits-per-byte (`wupz`) dense packing.
   - Implemented Metal GPU compute parity gates (`max_abs = 0.0 <= 1e-5` between reference Rust and GPU shader execution).
2. Zero-Float-Drift Arithmetic & Spectral Rigidity (`mem-16-10` & `cogito`):
   - Built exact rational phase accumulators over Z_{2^32} that run 9,000,000,000 steps with zero floating-point drift, alongside hardware `CLZ`-seeded integer square-root kernels (`1M` evaluations in `5.52 ms`).
3. Production Python-to-Systems Modernization:
   - Across fintech (Citadel / Morgan Stanley / Sixth Street) and AI SaaS (ASAPP), I have repeatedly replaced Python/JVM bottlenecks with high-concurrency Go and Rust services—the same philosophy behind PFN's Rustuna.

Selected Open-Source Repositories:
- https://github.com/8b-is/cogito (Zero-alloc Rust ternary bitplanes, exact rational lattices & ABI v1 witnesses)
- https://github.com/8b-is/zenUltraCooker (Rust workspace + Metal GPU parity kernel + native iOS/Android apps)
- https://github.com/peterlodri-sec (70+ systems, cryptography, and NixOS infrastructure repositories)

I am currently in Tokyo and excited to contribute to PFN's AI infrastructure and systems engineering teams.

Sincerely,
Péter József Lódri
peter.lodri@gmail.com | +36 20 391 5161
```