That their hearts may be comforted, they being knit together in love and unto all the riches of the full assurance of understanding, unto the full knowledge of the mystery of God, Christ, Praying at the same time for us also, that God would open to us a door for the word, to speak the mystery of Christ (because of which also I am bound),
Colossians 2:2: That their hearts may be comforted, they being knit together in love and unto all the riches of the full assurance of understanding, unto the full knowledge of the mystery of God, Christ,
Colossians 4:3: Praying at the same time for us also, that God would open to us a door for the word, to speak the mystery of Christ (because of which also I am bound),
Making known to us the mystery of His will according to His good pleasure, which He purposed in Himself, By which, in reading it, you can perceive my understanding in the mystery of Christ, Which in other generations was not made known to the sons of men, as it has now been revealed to His holy apostles and prophets in spirit,
Ephesians 1:9: Making known to us the mystery of His will according to His good pleasure, which He purposed in Himself,
Ephesians 3:4-5: By which, in reading it, you can perceive my understanding in the mystery of Christ, Which in other generations was not made known to the sons of men, as it has now been revealed to His holy apostles and prophets in spirit,
In the beginning was the Word, and the Word was with God, and the Word was God. And the Word became flesh and tabernacled among us (and we beheld His glory, glory as of the only Begotten from the Father), full of grace and reality.
John 1:1: In the beginning was the Word, and the Word was with God, and the Word was God.
John 1:14: And the Word became flesh and tabernacled among us (and we beheld His glory, glory as of the only Begotten from the Father), full of grace and reality.
John 1:1: 太初 有话/话存在 , 话与神同在, 话就是神。
John 1:14: 话成了肉体, 支搭帐幕在我们中间, 我们也见过祂的荣耀, 正是从父而来独生子的荣耀, 丰丰满满地有恩典,有实际。
In whom all the treasures of wisdom and knowledge are hidden. That the God of our Lord Jesus Christ, the Father of glory, may give to you a spirit of wisdom and revelation in the full knowledge of Him, O the depth of the riches and wisdom and knowledge of God! How unsearchable are His judgments and untraceable His ways!
Colossians 2:3: In whom all the treasures of wisdom and knowledge are hidden.
Ephesians 1:17: That the God of our Lord Jesus Christ, the Father of glory, may give to you a spirit of wisdom and revelation in the full knowledge of Him,
Romans 11:33: O the depth of the riches and wisdom and knowledge of God! How unsearchable are His judgments and untraceable His ways!
For in Him dwells all the fullness of the Godhead bodily, The grace of the Lord Jesus Christ and the love of God and the fellowship of the Holy Spirit be with you all.
Colossians 2:9: For in Him dwells all the fullness of the Godhead bodily.
2 Corinthians 13:14: The grace of the Lord Jesus Christ and the love of God and the fellowship of the Holy Spirit be with you all.
And you have been made full in Him, who is the Head of all rule and authority. To me, less than the least of all saints, was this grace given to announce to the Gentiles the unsearchable riches of Christ as the gospel
Colossians 2:10: And you have been made full in Him, who is the Head of all rule and authority.
Ephesians 3:8: To me, less than the least of all saints, was this grace given to announce to the Gentiles the unsearchable riches of Christ as the gospel
Hello, my name is Wei Sheng. I studied system analysis at graduate school in China, and I have been working in Japan as an IT engineer since 2016. My main language is Python, and I have experience from backend to cloud infrastructure.
In recent projects, I used FastAPI to build AI applications and wrote automated tests with pytest. I worked with SQLAlchemy for ORM design and PostgreSQL for database implementation. I also wrote shell scripts for automation and CI/CD pipelines.
I have hands-on experience with Generative AI and RAG. I built an internal agent system and monitoring automation system using these technologies. On the cloud side, I have worked with both AWS and Azure. On Azure, I handled VM setup, Blob storage, network, and ACR deployment.
I have worked in Japanese-speaking environments for more than eight years, so communication is not a problem. I believe I can contribute to this project with my skills and experience. Thank you.
想定質問と回答(英語)
Q1: What did you use FastAPI for? A1: I built REST APIs for AI applications. I documented them with Swagger and tested them with pytest.
Q2: Can you explain your experience with RAG? A2: I used LangChain and AWS Bedrock to build a RAG system. It connected a knowledge base with LLMs to provide accurate answers.
Q3: What Azure services have you worked with? A3: I set up Azure VM and Blob storage. I also used Azure Container Registry to deploy AI services.
逆質問(英語)
How will Generative AI and RAG be used in this project?
How large is the development team?
Is the schedule of on-site and remote days fixed each week?
The most relevant project is the “Intelligent Agent Platform” I developed. In this project, I used Python and FastAPI to build a serverless AI agent system. I managed data with SQLAlchemy and wrote automated tests with pytest. The database was PostgreSQL. I also wrote shell scripts for deployment and Terraform reviews.
For the AI part, I implemented a RAG system using LangChain, AWS Bedrock, and Kendra. This is directly related to Generative AI and RAG-based application development. In addition, I have experience with Azure VM and Blob storage setup, so I believe my background fits well with the requirements of this project.
Giving thanks to the Father, who has qualified you for a share of the allotted portion of the saints in the light; Who delivered us out of the authority of darkness and transferred us into the kingdom of the Son of His love,
Giving thanks to the Father , who has qualified you for a share of the allotted portion of the saints in the light; Who delivered us out of the authority of darkness and transferred us into the kingdom of the Son of His love,
And I have come down to deliver them out of the hand of the Egyptians and to bring them up out of that land to a good and spacious land, to a land flowing with milk and honey, to the place of the Canaanites and the Hittites and the Amorites and the Perizzites and the Hivites and the Jebusites. The next day he saw Jesus coming to him and said, Behold, the Lamb of God, who takes away the sin of the world! Truly, truly, I say to you, Unless the grain of wheat falls into the ground and dies, it abides alone; but if it dies, it bears much fruit.
And I have come down to deliver them out of the hand of the Egyptians and to bring them up out of that land to a good and spacious land to a land flowing with milk and honey to the place of the Canaanites and the Hittites and the Amorites and the Perizzites and the Hivites and the Jebusites The next day he saw Jesus coming to him and said Behold the Lamb of God who takes away the sin of the world Truly, truly, I say to you Unless the grain of wheat falls into the ground and dies it abides alone but if it dies it bears much fruit
As newborn babes, long for the guileless milk of the word in order that by it you may grow unto salvation, How sweet are Your words to my taste! Sweeter than honey to my mouth!
As newborn babes long for the guileless milk of the word in order that by it you may grow unto salvation How sweet are Your words to my taste Sweeter than honey to my mouth
As therefore you have received the Christ, Jesus the Lord, walk in Him, Having been rooted and being built up in Him, and being established in the faith even as you were taught, abounding in thanksgiving. For in Him dwells all the fullness of the Godhead bodily, And you have been made full in Him, who is the Head of all rule and authority.
As therefore you have received the Christ, Jesus the Lord walk in Him Having been rooted and being built up in Him and being established in the faith even as you were taught abounding in thanksgiving For in Him dwells all the fullness of the Godhead bodily And you have been made full in Him who is the Head of all rule and authority
Buried together with Him in baptism, in which also you were raised together with Him through the faith of the operation of God, who raised Him from the dead. And you, though dead in your offenses and in the uncircumcision of your flesh, He made alive together with Him, having forgiven us all our offenses; Wiping out the handwriting in ordinances, which was against us, which was contrary to us; and He has taken it out of the way, nailing it to the cross.
Buried together with Him in baptism in which also you were raised together with Him through the faith of the operation of God who raised Him from the dead And you, though dead in your offenses and in the uncircumcision of your flesh He made alive together with Him having forgiven us all our offenses Wiping out the handwriting in ordinances which was against us which was contrary to us and He has taken it out of the way nailing it to the cross
Having been rooted and being built up in Him, and being established in the faith even as you were taught, abounding in thanksgiving. Stripping off the rulers and the authorities, He made a display of them openly, triumphing over them in it.
Having been rooted and being built up in Him and being established in the faith even as you were taught abounding in thanksgiving Stripping off the rulers and the authorities He made a display of them openly triumphing over them in it
Even though He was a Son, learned obedience from the things which He suffered. And having been perfected, He became to all those who obey Him the source of eternal salvation, ​ through whom we have received grace and apostleship unto the obedience of faith among all the Gentiles on behalf of His name,
Therefore also God highly exalted Him and bestowed on Him the name which is above every name, that in the name of Yeshua every knee should bow, of those who are in heaven and on earth and under the earth, and every tongue should openly confess that Yeshua Messiah is Lord to the glory of God the Father.
Who, existing in the form of God, did not consider being equal with God a treasure to be grasped, You have heard that I said to you, I am going away and I am coming to you. If you loved Me, you would rejoice because I am going to the Father, for the Father is greater than I.
When Messiah our life is manifested, then you also will be manifested with Him in glory. I am crucified with Messiah; and it is no longer I who live, but it is Messiah who lives in me; and the life which I now live in the flesh I live in faith, the faith of the Son of God, who loved me and gave Himself up for me.
So then, my beloved, even as you have always obeyed, not as in my presence only but now much rather in my absence, work out your own salvation with fear and trembling; For it is God who operates in you both the willing and the working for His good pleasure. Do all things without murmurings and reasonings that you may be blameless and guileless, children of God without blemish in the midst of a crooked and perverted generation, among whom you shine as luminaries in the world, holding forth the word of life, so that I may have a boast in the day of Messiah that I did not run in vain nor labor in vain. (Philippians 2:12-16 Recovery Version)
Let this mind be in you, which was also in Messiah Yeshua, Who, existing in the form of God, did not consider being equal with God a treasure to be grasped, But emptied Himself, taking the form of a slave, becoming in the likeness of men; And being found in fashion as a man, He humbled Himself, becoming obedient even unto death, and that the death of a cross. Therefore also God highly exalted Him and bestowed on Him the name which is above every name, That in the name of Yeshua every knee should bow, of those who are in heaven and on earth and under the earth, And every tongue should openly confess that Yeshua Messiah is Lord to the glory of God the Father. (Philippians 2:5-11 Recovery Version)
English “For Azure only, Bicep is simpler and supports new features quickly. For multi-cloud or team operations, Terraform has more use cases and flexibility.”
Direct Connect は AWS と自社データセンターを専用線でつなぐサービスです。 Azure では ExpressRoute が対応します。 高速で安定した専用回線接続が可能です。
PrivateLink と PrivateLink Service は、AWS のサービスをプライベート IP で利用する仕組みです。 Azure では Private Endpoint と Private Link Service が同じ役割です。 クライアント側は Private Endpoint、サービス提供側は Private Link Service です。
最後に、VPC Endpoint です。 これは VPC 内から AWS サービスをプライベートアクセスする仕組みです。 Azure では Service Endpoint または Private Endpoint が対応します。 Service Endpoint は子ネットから Azure PaaS にアクセス。 Private Endpoint はリソースに専用のプライベート IP を割り当てて完全にプライベートでアクセスできます。
Hello, my name is Andrew. I studied Electrical Engineering in undergraduate school and System Analysis in graduate school, and since 2016 I have been working in Japan as a software engineer. I have more than eight years of experience in cloud, AI, and automation projects.
Recently, I worked on building an internal agent platform using AWS and Azure. I implemented AI services, used RAG for knowledge search, and also considered security, such as access control and monitoring. I have long experience with Linux and Python. I also have some background in Firewall and network monitoring.
I like learning new technologies. I am also interested in Bicep and IaC tools. I believe I can contribute to this project with my cloud and AI experience, and I am motivated to grow as a cloud engineer. Thank you very much.
■英文 Anticipated Questions & Answers
Q1. Do you have experience with Azure? A1. Yes. I mainly worked with AWS recently, but I also used Azure Portal to implement AI services. I am willing to learn more Azure services quickly.
Q2. How about your Firewall or security experience? A2. I configured monitoring and firewall rules in cloud environments. I also analyzed logs for security. I understand the basics and can handle it.
Q3. How much Python experience do you have? A3. I have more than eight years of experience. I used it for API development, data processing, and AI implementation. Functions and testing are no problem.
■英文 My Questions to Them
For the hybrid work style, do you have a fixed day to go to the office?
Which Azure AI services will be mainly used in this project?
Regarding security, will the main focus be firewall setup, or also monitoring and operation?
For example, in June–July 2025, I worked on an Internal Agent Platform project. I used AWS Bedrock and Kendra, and implemented RAG for AI services. I also handled security design and Infrastructure as Code with Terraform. Python was the main language. Through this project, I gained experience in implementing AI services in the cloud. I believe I can apply the same skills to Azure, using the portal and managed AI services.
Hello, my name is Andrew. I studied Electrical Engineering and System Analysis in China, and since 2016 I have been working in Japan as a software engineer. I mainly work with Python, and I have experience from backend to frontend, cloud, and AI. In recent years, I focused on Generative AI and RAG systems. I designed and built internal agent platforms and monitoring automation systems using LLMs. I also have hands-on experience with FastAPI, pytest, and SQLAlchemy. I have used PostgreSQL in production, and I write Shell scripts for automation. On the infrastructure side, I have worked with Azure VM, ACR, Blob Storage, and also AWS services. As a project leader, I managed tasks from design to testing and deployment. I often used GitHub and Swagger for documentation and team collaboration. I have been working in a Japanese-speaking environment for over 8 years, so communication is not a problem. I would like to contribute my skills and experience to your AI application development project. Thank you.
Possible Questions & Answers
Q1. Can you explain your Python experience? A1. I have used Python for more than 8 years. I developed APIs with FastAPI, automated tests with pytest, and designed database models with SQLAlchemy. Recently, I worked on RAG and Generative AI systems.
Q2. What experience do you have with Azure? A2. I built systems on Azure VM, used ACR for container images, and worked with Blob Storage and networking. I also have AWS and GCP experience.
Q3. What was your role in the team? A3. I often took the role of a technical leader. I did system design, task assignment, and also coding and testing myself.
English At MonotaRO, I developed a product recommendation API. I implemented it using Python and FastAPI, and managed data with SQLAlchemy. The database was PostgreSQL. I automated tests with pytest and integrated them into the CI/CD pipeline. I was responsible for the entire process, from design to implementation, testing, and deployment.
English At SynX, I built an internal agent platform. I used Python for a serverless setup and integrated RAG with a knowledge base. The system provides automated responses using Generative AI. I used AWS Bedrock and Kendra, but the same structure can be applied on Azure. I documented with Markdown and Swagger, managed the code on GitHub, and automated infrastructure using Terraform.
English Another project at SynX is called “Walkure Operator,” an intelligent monitoring automation system. I integrated Zabbix with multiple LLMs using Python. I used LangChain and RAG to enable knowledge search. Documentation was written in Markdown and code managed with GitHub. The system automated alert summarization, root cause estimation, and predictive maintenance.
English (concise) This is an internal agent platform. I built it serverless with Python. It integrates RAG and a knowledge base to provide automated responses using LLMs. Used AWS Lambda, API Gateway, DynamoDB. Managed infrastructure with Terraform and version control with GitHub. Tests implemented using pytest. Generative AI use cases are included.
English (concise) This is an intelligent monitoring automation system. Integrated Zabbix with multiple LLMs using Python. Supports alert summarization, root cause estimation, predictive maintenance. Used LangChain and RAG for knowledge search. MCP protocol ensures LLM compatibility. Documentation in Markdown, version control via GitHub.
English (concise) This system analyzes Backlog project data with AI and automatically generates HTML slides. Includes voice narration and interactive visualization. Implemented AI analysis and visualization in Python. Published and maintained on GitHub.
English (concise) Built internal RAG system in Python using AWS generative AI sample projects. Integrated LLMs with Amazon Bedrock and Kendra. Managed infrastructure and automated reviews with Terraform and GitHub Actions.
Hello, my name is Andrew. I am an experienced software engineer with over eight years of professional experience in Japan, working entirely in Japanese-speaking environments. I graduated with a master’s degree in Electrical Engineering and System Analysis from Zhejiang University in China, and since coming to Japan, I have been involved in a wide range of projects, including web systems, logistics automation, image processing, big data pipelines, and AI-driven platforms.
Most recently, I have been developing Python-based batch and automation systems, including AI-powered operational monitoring systems and internal agent platforms. I have designed, implemented, and maintained backend services, data pipelines, and batch jobs, often integrating them with BI platforms and cloud services such as AWS Lambda, API Gateway, and DynamoDB. I also have experience leading projects, coordinating teams, and preparing documentation and procedures.
I am particularly skilled in Python for batch processing, asynchronous programming, and system integration. I am proactive, collaborative, and able to communicate effectively with team members and stakeholders. I am excited about the opportunity to contribute my expertise to your Python batch development and BI integration projects, and I am confident I can help strengthen the stability and efficiency of your systems.
面试可能问你的问题及示例回答
Can you describe your experience with Python batch development? Answer: I have developed and maintained multiple Python batch systems for data processing, monitoring, and automation. For example, at my current company, I built automated workflows that process project data, integrate with BI platforms, and generate reports, ensuring stable operation and handling exceptions effectively.
Have you worked with JP1, Shell, or C in batch systems? Answer: Yes, I have experience with Shell scripting for automation, and I have developed some legacy system integrations using C. In most cases, I combine Python with these tools to create reliable batch processes and data pipelines.
How do you ensure communication within a team, especially when acting as a lead? Answer: I maintain clear documentation, hold regular meetings, and proactively check on task progress. I make sure everyone understands their responsibilities and feel comfortable sharing concerns. This approach has helped me manage teams efficiently and deliver projects on time.
Have you migrated batch systems to BI platforms? Answer: Yes. For instance, I have migrated data processing pipelines into BI systems, automating reporting and dashboards. I focus on preserving data integrity, ensuring smooth ETL processes, and creating clear operational procedures.
Do you have experience with AWS services? Answer: I have worked with AWS Lambda, API Gateway, DynamoDB, and S3 in serverless architectures. I have integrated batch processes and internal automation systems on AWS, though my primary focus is on Python development.
你可以问面试官的问题
Can you tell me more about the current Python batch projects and their main challenges?
How is the team structured, and what is the expected collaboration between the lead and SE roles?
What BI platforms and tools are currently used, and are there any planned migrations or upgrades?
How often are the batch jobs executed, and what measures are in place to ensure stability and monitoring?
Are there opportunities to contribute to process improvement or automation initiatives within the team?
My name is Andrew. I studied system analysis at Zhejiang University and have been working in Japan as an engineer since 2016. Over the past eight years, I have gained experience in full-stack development, cloud platforms, and more recently, AI-driven automation using large language models. In particular, I have designed and implemented RAG-based systems, optimized prompt engineering, and integrated LLMs with enterprise applications. My main language is Python, but I am also comfortable with JavaScript, Golang, and cloud services such as AWS and GCP. I always try to turn new technologies into practical solutions. I would like to contribute to your AI solutions in your industry with both my technical skills and implementation experience. Thank you.
Project introduction (2)
The first project is the “Intelligent Agent Platform.” I built an internal agent system using AWS Bedrock and Kendra, combined with RAG architecture. It enabled knowledge search and automated Terraform review, which greatly improved productivity.
The second project is “Walkure Operator,” an intelligent monitoring system. I integrated Zabbix with multiple LLMs to summarize alerts, estimate root causes, and even predict failures. By using LangChain and vector databases, I improved search accuracy and reduced manual operation workload.
Likely questions and answers (English)
Q1: Why do you want to work in the financial industry? A: Because the financial industry deals with massive data and requires high reliability. I believe AI solutions such as RAG and LLMs can bring significant efficiency and new value there.
Q2: What experience do you have with prompt engineering? A: In the monitoring automation system, I tested and optimized multiple prompt templates to improve accuracy in summarizing alerts and predicting causes. This reduced errors and improved operational efficiency.
Q3: What role do you usually play in a team? A: I often take the role of project leader, handling requirements, design, development, and deployment. At the same time, I contribute directly to coding and implementation.
Questions to ask them (English)
Which areas of AI solutions in the financial sector are you focusing on the most right now?
How do you evaluate the quality and reliability of RAG or LLM-based systems in your company?
Could you tell me more about your team structure and development process?
Hello, my name is Wei Sheng. I completed my Master’s degree in System Analysis at Zhejiang University in China, and since 2016, I have been working in Japan in a fully Japanese business environment. Over the past eight years, I have gained wide-ranging experience in backend and frontend development, cloud infrastructure, automation, and AI integration, while also leading multiple projects as a technical leader.
I have strong hands-on experience with AWS services, including API Gateway, Lambda, ECS, and CloudWatch, building serverless systems and container-based solutions. I am also proficient in Python, Go, and JavaScript, and I have worked on real-time communication and data processing systems.
While I have not yet developed a large-scale product directly with WebRTC, I have studied its architecture, and I have experience with audio streaming, codecs, and Docker-based development. Combined with my background in Linux environments and monitoring dashboards, I believe I can quickly get up to speed and contribute to your project.
In addition, I am fluent in Japanese, Chinese, and English, which allows me to work smoothly in multicultural teams. I am very motivated to contribute to your WebRTC-based cloud service for multi-connection audio streaming, ensuring quality and scalability. Thank you very much.
英語 Expected Questions & Answers
Q1: Do you have direct experience with WebRTC? A1: Not directly in a production system, but I have worked on real-time communication and streaming with Go and Python, and I have studied WebRTC and audio codecs. I am confident I can adapt quickly.
Q2: What AWS services have you used? A2: I have implemented serverless systems using Lambda, API Gateway, and DynamoDB, containerized applications on ECS, and monitoring with CloudWatch, including dashboards for metrics and logs.
Q3: How do you usually contribute to team projects? A3: I take responsibility for end-to-end processes, from design to deployment, and as a project leader, I manage schedules and make technical decisions. I also communicate effectively across teams and cultures.
英語 You can ask them
Which WebRTC framework or OSS are you currently using, for example LiveKit or others?
What are the biggest technical challenges in your audio streaming system?
Could you tell me about the team size and the balance between remote work and office work?
コーデックというのは「コーダー」と「デコーダー」の意味で、音声を**圧縮して送る技術**のことです。 インターネットで会話するときは、そのまま送るとデータが大きすぎるので、コーデックで小さくしてから送ります。 代表的なのは Opus というコーデックで、高音質で遅延も少なく、WebRTCではよく使われています。
Hello, my name is Wei Sheng. I studied Electrical Engineering and System Analysis at Zhejiang University in China, and since moving to Japan in 2016, I have been working in a fully Japanese environment as a software engineer. I have extensive experience in backend, frontend, cloud services, data infrastructure, and AI systems, and I have led multiple project teams as a technical leader.
Recently, I have worked on serverless systems using AWS and automation monitoring systems integrating large language models. I am proficient in Python, Go, and JavaScript, and I have basic knowledge of audio data processing and real-time communication. I am also fluent in English and Chinese, which allows me to collaborate effectively in multicultural teams.
I am excited about the opportunity to contribute my skills to your WebRTC-based multi-connection audio transmission cloud service, ensuring high-quality and scalable system development. Thank you for your time.
Q1: Do you have experience with WebRTC? A1: I haven’t worked with WebRTC directly, but I have experience developing real-time communication systems in Python and Go, and I can quickly get up to speed with WebRTC technology.
Q2: What is your experience with serverless architecture on AWS? A2: I have designed and implemented serverless systems using AWS Lambda, API Gateway, DynamoDB, and ECS, integrating automation and monitoring features.
Q3: How do you contribute to team projects? A3: As a project leader, I manage design, implementation, testing, deployment, and operation, making technical decisions and ensuring project progress and quality.
Could you tell me about the team size and structure for this project?
What are the key challenges in real-time audio processing that the team focuses on?
What is the balance between remote work and office presence?
今回の技術課題では、「Intelligent Presenter for HTML Slides」というプロダクトを開発し、BacklogのプロジェクトデータからAIを活用してHTMLスライドを自動生成するシステムを構築しました。MCP協議の採用、Vue 3とSlidevによるモダンなフロントエンド、そして日本語TTS機能まで統合した包括的なソリューションに仕上げました。この課題を通じて、Nulabの技術方針やプロダクト理念に深く共感し、特にBacklog AI助手の機能拡張という方向性で、既存サービスとの統合を意識した設計を心がけました。
Intelligent Presenter for HTML Slidesは、BacklogプロジェクトデータからAIを活用してHTMLスライドを自動生成するシステムです。
主要機能
BacklogプロジェクトデータをMCP協議で取得
AIによる10種類のテーマ別スライド自動生成
日本語TTS音声解説機能
リアルタイム流式処理によるスライド配信
解説文
こんにちは。私が開発したIntelligent Presenter for HTML Slidesについてご紹介させていただきます。このシステムは、BacklogのプロジェクトデータからAIを使ってHTMLスライドを自動で生成するシステムです。Backlogプロジェクトのデータを自動で取得し、AIが十種類のテーマに分けてスライドを作成します。さらに、日本語の音声解説も付けることができます。
今回の技術課題では、「Intelligent Presenter for HTML Slides」というプロダクトを開発し、BacklogのプロジェクトデータからAIを活用してHTMLスライドを自動生成するシステムを構築しました。MCP協議の採用、Vue 3とSlidevによるモダンなフロントエンド、そして日本語TTS機能まで統合した包括的なソリューションに仕上げました。この課題を通じて、Nulabの技術方針やプロダクト理念に深く共感し、特にBacklog AI助手の機能拡張という方向性で、既存サービスとの統合を意識した設計を心がけました。
特に最近では、生成AIとLLMを活用したシステム開発に力を入れております。現職のSynXでは、ZabbixとMistral、Mixtralなどの複数のLLMを統合する監視自動化システム「Walkure Operator」のアーキテクト兼リードデベロッパーとして、アラート要約や原因推定、予知保全システムを設計・実装いたしました。また、AWSのGenerative AI Use Casesフレームワークを活用した社内エージェントシステムの構築も手がけ、Model Context Protocolを採用したサーバーレス構成による自動化システムを開発しております。
A2: 最も代表的なプロジェクトは「Walkure Operator」です。ZabbixとMistral、Mixtralなどの複数のLLMを統合し、監視アラートの要約、根本原因の推定、予知保全を自動化するシステムを設計・開発いたしました。Model Context Protocolを採用してLLMの互換性を確保し、LangChainやRAG構成による知識検索機能も実装しております。また、AWSのGenerative AI Use Casesフレームワークを活用した社内エージェントシステムでは、Amazon BedrockやKendraと連携したサーバーレス構成を構築し、Terraform自動レビュー機能の評価も行いました。