Tatsuki Okada
エンジニア @ シェルパ・アンド・カンパニー株式会社 / Cierpa & Co.
About
[Profile Summary]I work at the intersection of NLP, large language models (LLMs), multimodal AI, and production engineering. I am conducting graduate research at NAIST's Natural Language Processing Laboratory (Watanabe Lab), while working on Vision-Language Model (VLM) R&D at Cierpa & Co.At ELYZA, Inc., I have spent more than four years building enterprise LLM, RAG, and agent systems across requirements, implementation, evaluation, deployment, and operations. I continue to work with ELYZA on practical GenAI systems. I have also gained backend and cloud engineering experience at DeNA, CyberAgent, CARTA HOLDINGS, Nikkei, and COLOPL using Go, Python, GCP, AWS, and GKE.My research includes correctness-guaranteed synthetic data generation for Japanese LLM reasoning and controlled evaluation of encoder- and decoder-based sentence embeddings. I am especially interested in connecting models, evaluation data, inference infrastructure, and applications into AI systems that work reliably in practice.[Experience]Cierpa & Co.VLM-centered R&D for AI products in the sustainability and ESG domain, applying experience in LLMs, multimodal models, evaluation design, and data preparation.ELYZA, Inc.Built enterprise LLM, RAG, and agent systems across requirements, implementation, evaluation, deployment, and operations. Worked on domain-adapted LLM fine-tuning, production deployment, evaluation infrastructure, and shared internal libraries. I continue to work with ELYZA.Other internshipsBackend and cloud engineering at DeNA, CyberAgent, CARTA HOLDINGS, Nikkei, and COLOPL: Go APIs, Kubernetes / GKE, MySQL optimization, RAG demos with Python / FastAPI / React, and RPC servers for large-scale game services.[Education]NAIST, Natural Language Processing Laboratory (Watanabe Lab)[Research & Publications]NLP 2026: Performance analysis of sentence embeddings with encoder / decoder architectures.NLP 2025: Improving Japanese LLM reasoning through correctness-guaranteed synthetic data generation.[Skills]Go / Python / TypeScriptGCP / AWS / Kubernetes / GKE / DockerPostgreSQL / MySQL / Database tuningNLP / LLM / VLM / RAG / Agents / Evaluation / Machine Learning[Additional Information]I also contribute to the open-source community through model and dataset releases on Hugging Face.
Japan
Kanagawa
Information Technology & Services
Hugging Face製品, PyTorch, VLM, 人工知能 (AI), PostgreSQL, Google Cloud Platform (GCP), Kubernetes, Google Kubernetes Engine (GKE), アマゾンウェブサービス (AWS), MySQL, Nginx, Webアプリケーション開発, TypeScript, LLM, Natural Language Processing, ウェブ開発, 自然言語処理, RPC, Go (プログラミング言語), JavaScript
Experience

エンジニア
東京都, 日本
I focus on Vision-Language Model (VLM) research and development for AI products in the sustainability and ESG domain. I apply my experience in LLMs, multimodal models, evaluation design, and data preparation to improve and validate models for practical use cases. --- サステナビリティ・ESG 領域の AI プロダクト開発で、VLM (Vision-Language Model) を中心とした研究開発に従事しています。LLM、マルチモーダルモデル、評価設計、データ整備の知見を活かし、実用を見据えたモデル改善と検証に取り組んでいます。 Ref: https://cierpa.co.jp/posts/20260604

AI Engineer
東京都, 日本
I worked on enterprise LLM, RAG, and agent systems across requirements, implementation, evaluation, deployment, and operations. I contributed to domain-adapted LLM fine-tuning, evaluation, and production deployment for industries such as insurance, recruiting, finance, and legal, and served as a main engineer on multiple projects for enterprise and public-sector clients. On the R&D side, I worked on inference acceleration, correctness-guaranteed synthetic data generation, and evaluation datasets and metrics. I also released research artifacts on Hugging Face, connecting production engineering and research to improve LLM quality. --- 企業向けの LLM / RAG / Agent システムを、要件整理から実装、評価、本番運用まで担当しました。保険、求人、金融、法務などの領域で、ドメイン特化 LLM のファインチューニング、評価、本番デプロイに取り組み、大手企業・省庁を含む複数案件で主担当を経験しました。 R&D では、推論の高速化、正解保証付き合成データ生成、評価データセット・評価指標の整備に取り組みました。研究成果のモデルやデータセットを Hugging Face で公開し、実務と研究の両面から LLM の品質改善に関わりました。 Ref: https://www.anlp.jp/proceedings/annual_meeting/2025/pdf_dir/P5-8.pdf, https://huggingface.co/datasets/elyza/JaMARD

ソフトウェアエンジニア
東京都, 日本
I extended QuickMTG, a Slack-based meeting-room reservation application, using Go. I implemented priority rules and floor-level defaults, improved the reservation algorithm, and supported post-release QA. --- Go を用いて、Slack 会議室予約アプリ「QuickMTG」の機能拡張を担当しました。優先度設定やフロアごとのデフォルト設定を実装し、予約アルゴリズムの改善とリリース後の QA 対応まで行いました。 Ref: https://engineering.dena.com/blog/2024/12/intern_report_quickmtg/

ソフトウェアエンジニア
東京都, 日本
I worked on Go backend API development and container infrastructure using Kubernetes and Google Kubernetes Engine (GKE). The internship gave me experience across both application development and cloud infrastructure for reliable service delivery. --- Go によるバックエンド API 開発と、Kubernetes / GKE を用いたコンテナ基盤の構築に取り組みました。アプリケーション実装とクラウドインフラの両面から、サービスを安定して動かすための開発を経験しました。

ソフトウェアエンジニア
東京都, 日本
I also worked on bottleneck analysis, performance improvements, and new feature development for a web-crawling application. Through these projects, I gained hands-on backend engineering experience with Go, GCP, and Nginx in production-oriented environments. --- Web クローリングアプリのボトルネック調査、性能改善、新機能実装にも取り組みました。GCP、Nginx、Go を用いて、実運用を意識したバックエンド開発を経験しました。

ソフトウェアエンジニア
東京都, 日本
I developed backend services using Go and AWS. I worked on MySQL schema design, query optimization, and infrastructure improvements, gaining practical experience in building systems with database performance and maintainability in mind. --- Go と AWS を用いたバックエンドサービス開発に取り組みました。MySQL のスキーマ設計、クエリ最適化、インフラ最適化を担当し、データベース性能と運用しやすさを意識した実装を経験しました。

ソフトウェアエンジニア
東京都, 日本
I designed and implemented a RAG system over Nikkei article data. Using Python, FastAPI, and React, I built a demo experience that combined article retrieval, question answering, and frontend presentation for news subscription use cases. --- 日経記事データを対象とした RAG システムを設計・実装しました。Python、FastAPI、React を用いて、記事検索、質問応答、フロントエンド表示を含む新聞購読体験のデモを開発しました。

ソフトウェアエンジニア
東京都, 日本
I implemented RPC servers for large-scale game services. I used Go to build server-side functionality as part of game infrastructure, gaining backend development experience in systems that support high-traffic services. --- 大規模ゲームサービス向けの RPC サーバー実装に取り組みました。Go を用いて、ゲームインフラの一部となるサーバーサイド機能を開発し、大規模サービスを支えるバックエンド開発を経験しました。
Tatsuki Okada's Contact Information
Phone
Find the Right Leads
Find Verified Contact Data
What LeadContact does well
Find verified emails, phone numbers, and decision-makers with 98% accuracy.
Find Leads
Find the right people by company, role, industry, location, and more.
925M+ professional profiles

Find Emails
Access verified email addresses for your target contacts.
657M+ emails

Find Phone Numbers
Get cross-validated phone data from multiple top sources.
239M+ phone numbers

More Accurate. Lower Cost.
Find contact data in 1 tool with 98% accuracy
LeadContact integrates leading enrichment tools to deliver more accurate contact data—without paying for each one.
Great conversations start with the right contact.
It’s time to find yours.



