关于
Abel Wang (王静玺) has 20 years of experience in IT strategic planning, Architecture design…
Abel的文章
动态
2198 位关注者
-
Abel Wang 分享了此动态工作需要，最经常在上海，江南三月，目光所及皆是繁花与新绿，身边来往鲜衣怒马和高楼大厦。驻足路边想到一些，操记录，备忘 多年来，银行业对于科技的持续投入和积极探索，积累了大量的经验，并上升为信息化的理论体系，同时积累了大量的资产。很多重要的理念和方法广泛适用于多种行业，也看到这种推广和迁移在制造业、零售等领域的成功。在以下多个方面产生了参考设计模式 • 现代化核心系统，围绕产品、客户、合约、业务活动及账务、为业务发展提供了全局统一的轻量数据目录索引和服务支撑。正真实现了数据的纲举目张 • 强一致性的交易保障，为关键业务处理逻辑提供了处理范式 • 模型方法全面落地，推动了先进经验到共识的平稳过渡和快速推广 • 无感金融推动了跨行业、跨机构“可信可交换信息摘要”的标准普及 • 多层次数据模型分类体系（9大基础域+11类管理分析域） • 异步+事件的交易核心底层框架以异常事件处理 • 大规模信创及应对生产安全的设计补偿 • 主机系统的迁移 • 其它
-
Abel Wang 发布了此动态在新经济环境下，金融应突出其“经济基础设施”的特性，促进产业和区域经济的高质量发展才是首要目标。避免追求自身的高收益和高速发展．而不顾及服务对象的经营滞涨。在两个方面需要进一步突破：* “跨机构的可交换基础信息摘要”，商业机构之间有效的合作离不开“少而够用”的信息交换，通过建章立制，实现“机构/ 个人基本资料”“产品摘要”“合约基本资料”在可管理的前提下，有序交换。推动数宇凭证、可信商业网络相关技术的落地普及：* “OpenAPI”的进一步落实，实现在企业内业务处理层面使用到来白金融机构的“Open API”是金融服务 无感融入企业的经营。从长远来看“产业金融”会成为金融赋能经济的领先实践
-
Abel Wang 分享了此动态W wAbel Wang 分享了此动态Goldwind finishes the fruitful trip from AWEA, and goes on a new journey of its globalization. https://lnkd.in/eZ22CaD
-
Abel Wang 分享了此动态The summary of “The path to next-gen connected business(e-commerce)“. The key competency is going as following
-
Abel Wang 分享了此动态MaaC, messenger as a container, us getting to be a transaction platform, it start a new technology domain to run internet business
-
Abel Wang 赞了此动态Abel Wang 赞了此动态I asked 100 leaders what their biggest challenge is right now. A year ago, it was, "how do I get my team to use AI?" Now it's, "How do I get my team to STOP using AI SO BADLY?" Things I heard repeatedly: - "Everyone is generating a million garbage AI documents and no one reads them." - "My high performers are too busy to learn a totally different way of working, and my low performers use AI badly. There's no middle." - "AI has created more items to review, at a faster pace, rather than creating promised efficiency." This is a hard problem to fix because it's not a tool problem. AI amplifies whatever is in your head. If your people think clearly, AI is a jetpack. If they don't, well, AI is still a jetpack, but, the in the most chaotic sense imaginable. That's why most AI enablement is doomed. You can't solve it by teaching tools. You have to teach people to think; ironic, because so many people are using AI as a substitute for thinking. Talented people don't want to work at internal slop factories. If you don't get ahead of the chaos, performance becomes a race to the bottom: more pressure to act, less pressure to think. This is an obvious doom loop. That's why my programs all start with thinking. Week one of Couch to 5K for AI is about one thing: learning to tell the difference between mediocre and excellent work. How are you combatting bad AI use?
-
Abel Wang 赞了此动态Abel Wang 赞了此动态I have reviewed dozens of AI agent architectures this year alone. The same expensive, avoidable mistake keeps showing up every time… Teams are not failing because they chose the wrong model. They are failing because they chose the wrong pattern for the wrong problem. And nobody catches it until the system is already in production. Here is what I keep seeing on the ground: Engineers get excited about multi-agent systems rightfully so. They spin up parallel pipelines, build hierarchical orchestrators, stack composite patterns on top of each other. It looks impressive in a diagram. It looks very different on an infrastructure bill. The question nobody asks early enough is simple: Does this problem actually need this pattern? 📌 6 patterns, here is the cost vs. value reality of each: → Sequential: Lowest overhead of all. Tasks flow step by step in order. Start here and only move up when this pattern genuinely cannot do the job. → Parallel: Compresses time but multiplies cost. Only worth deploying when tasks are truly independent and latency is your real bottleneck. → Hierarchical: Best long-term ROI when sub-agents are reusable across workflows. The orchestration layer pays for itself only at scale. → Generator-Critic: One agent creates, another reviews, loops repeat until quality passes. Only reach for this when output quality is the primary success metric. → Human-in-the-Loop: Latency here is intentional, not a bug. Non-negotiable for compliance, billing, and any decision with real consequences. → Composite: Most expensive to run. Reserve strictly for end-to-end workflows where the full pattern stack genuinely earns its cost. 📌 What most architects are missing in 2026: Design patterns alone are not enough anymore. The protocol stack underneath has changed everything: → MCP handles how your agents connect to tools → A2A handles how your agents coordinate with each other → Both now governed by Linux Foundation, backed by Google, Anthropic, OpenAI, Microsoft, and AWS This is not experimental anymore. This is the new production baseline. The teams winning with AI agents in 2026 are not the ones with the most sophisticated diagrams. They are the ones who understand exactly what they are building and exactly what it costs. 📌 If you want to go deeper on AI Agents, my free newsletter breaks down everything you need to know: https://lnkd.in/esJDdA5Q Save 💾 → React 👍 → Share ♻️ Follow for more on AI Agents, architecture, and what actually holds up in production. #AIAgents #MultiAgentSystems #AgenticAI #AIArchitecture #EnterpriseAI
-
Abel Wang 赞了此动态Abel Wang 赞了此动态Product leaders, stop hiding behind docs! If your team is still spending all their time in PRDs and product strategy docs, they're not operating in 2025. AI prototyping has literally changed the game. Here's how teams should do it: — THE OLD WAY (STILL HAUNTS MOST ORGS) 1. Ideation (~5% actually prototyped) “We should build X.” Cool idea. But no prototype. Just a Notion doc and crossed fingers. 2. Planning (~15% use real prototypes) Sketches in Figma. Maybe a flowchart. But nothing a user could actually click. 3. Discovery (~50% try protos) Sometimes skipped. Sometimes just a survey. Rarely ever tested with something interactive. 4. PM Handoff (~5%) PM: “Here’s the PRD.” Design: “Uhh… where’s the prototype?” PRDs get passed around like homework. 5. Design Design scrambles to build something semi-clickable, just so people stop asking “what’s the plan?” 6. Eng Start Engineering starts cold. No head start. They’re building from scratch because nothing usable exists. — WHAT HAPPENS - Loop after loop. Everyone frustrated. - Slow launches. Lots of guesswork. - And no one truly understands the user until it’s too late. — THE NEW WAY (THIS IS HOW WINNERS SHIP) 1. Ideation PMs don’t just write ideas. They prototype them. Want to solve a user problem? Click, drag, test. There. No waiting. No “someday.” You build it, even if it’s ugly. 2. Planning Prototypes are the roadmap. You walk into planning with a live flow, not a list of features. And everyone’s like: “Oh. THAT’S what you meant.” 3. Discovery Real users. Real prototypes. You send them a flow and you watch them break it. You’re not guessing anymore. You’re observing. 4. PM Handoff PMs don’t just hand off docs. They ship working demos alongside the PRD. No more “interpret this paragraph.” Just click and see it work. 5. Design Designers don’t start from scratch. They take what’s already tested, validated, and tweak it. Suddenly, “design time” is “refinement time.” 6. Eng Start Engineers don’t wait around. They start with something usable. If not, they prompt an AI tool to build it. And we’re off to the races. — If you want to see how AI prototyping actually works (and learn from expert Colin Matthews), check out the deep dive: https://lnkd.in/eJujDhBV
-
Abel Wang 赞了此动态Abel Wang 赞了此动态𝐇𝐨𝐰 𝐭𝐨 𝐚𝐧𝐬𝐰𝐞𝐫 𝐚𝐧𝐲 𝐬𝐲𝐬𝐭𝐞𝐦 𝐝𝐞𝐬𝐢𝐠𝐧 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧 𝐮𝐬𝐢𝐧𝐠 𝗦𝘆𝘀𝘁𝗲𝗺 𝗗𝗲𝘀𝗶𝗴𝗻 𝗠𝗮𝘀𝘁𝗲𝗿 𝗧𝗲𝗺𝗽𝗹𝗮𝘁𝗲? Here is a master template that I used to discuss many #systemdesign problems in #interviews. 1. 𝐋𝐨𝐚𝐝 𝐁𝐚𝐥𝐚𝐧𝐜𝐞𝐫: Distributes incoming network traffic across multiple servers to ensure no single server bears too much demand. This helps increase the availability and reliability of applications. 2. 𝐀𝐏𝐈 𝐆𝐚𝐭𝐞𝐰𝐚𝐲: Acts as a gatekeeper for APIs, handling request routing, composition, and protocol translation. Often includes functionalities like authentication, monitoring, and load balancing. 3. 𝐒𝐭𝐚𝐭𝐢𝐜 𝐂𝐨𝐧𝐭𝐞𝐧𝐭 & 𝐂𝐃𝐍: Delivery of static assets (like images and scripts) that don't change often, using a Content Delivery Network (CDN) to speed up their delivery by caching content in multiple locations around the world. 4. 𝐌𝐞𝐭𝐚𝐝𝐚𝐭𝐚 𝐒𝐞𝐫𝐯𝐞𝐫 𝐚𝐧𝐝 𝐁𝐥𝐨𝐜𝐤 𝐒𝐞𝐫𝐯𝐞𝐫: Components that manage metadata and data blocks, respectively. Metadata servers store information about where data blocks are located and their properties, while block servers actually store the data chunks. 5. 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐞𝐝 𝐅𝐢𝐥𝐞 𝐒𝐭𝐨𝐫𝐚𝐠𝐞: Storage system designed to store data across multiple points of presence to increase data redundancy and reliability. 6. 𝐅𝐞𝐞𝐝 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧 𝐒𝐞𝐫𝐯𝐢𝐜𝐞 & 𝐐𝐮𝐞𝐮𝐞: Handles the processing and queuing of data needed to generate user-specific content feeds, like those seen on social media platforms. 7. 𝐒𝐡𝐚𝐫𝐝 𝐌𝐚𝐧𝐚𝐠𝐞𝐫 & 𝐃𝐢𝐫𝐞𝐜𝐭𝐨𝐫𝐲 𝐁𝐚𝐬𝐞𝐝 𝐏𝐚𝐫𝐭𝐢𝐭𝐢𝐨𝐧𝐢𝐧𝐠: Manages data sharding and partitioning strategies, optimizing data distribution and access by dividing larger databases into more manageable pieces. 8. 𝐍𝐨𝐭𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧 𝐒𝐞𝐫𝐯𝐢𝐜𝐞 & 𝐐𝐮𝐞𝐮𝐞: Manages sending notifications to users, handling both the queueing of notification messages and their delivery. 9. 𝐂𝐚𝐜𝐡𝐞 (𝐑𝐞𝐝𝐢𝐬/𝐌𝐞𝐦𝐜𝐚𝐜𝐡𝐞𝐝): In-memory data storage used to reduce the number of times data is read from slower databases, enhancing performance by serving repeated requests for the same data quickly. 10. 𝐕𝐢𝐝𝐞𝐨 𝐏𝐫𝐨𝐜𝐞𝐬𝐬𝐢𝐧𝐠 𝐐𝐮𝐞𝐮𝐞 & 𝐖𝐨𝐫𝐤𝐞𝐫𝐬: Dedicated system components for video processing tasks, queuing them and distributing the workload among multiple workers to process the videos as needed. 11. 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐞𝐝 𝐋𝐨𝐠𝐠𝐢𝐧𝐠 & 𝐓𝐫𝐚𝐜𝐢𝐧𝐠: Systems for logging and tracing the behavior of distributed systems, essential for debugging and monitoring the performance of distributed applications. 12. 𝐃���𝐭𝐚 𝐏𝐫𝐨𝐜𝐞𝐬𝐬𝐢𝐧𝐠 𝐒𝐲𝐬𝐭𝐞𝐦𝐬 (𝐇𝐚𝐝𝐨𝐨𝐩/𝐌𝐚𝐩𝐑𝐞𝐝𝐮𝐜𝐞, 𝐒𝐩𝐚𝐫𝐤): Frameworks used for processing large data sets with a distributed algorithm on a cluster. Ref: ✅𝗚𝗿𝗼𝗸𝗸𝗶𝗻𝗴 𝘁𝗵𝗲 𝗦𝘆𝘀𝘁𝗲𝗺 𝗗𝗲𝘀𝗶𝗴𝗻 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 - https://lnkd.in/g4Wii9r7 📌 Recommend DesignGurus.io's courses to earn 20% on each referral: https://lnkd.in/gz4rqr5k
-
Abel Wang 赞了此动态Abel Wang 赞了此动态From Silos to Synergy: 9 Best Practices for Integrating #DataArchitecture to Drive Digital Transformation with Case Studies Integrating data architecture is essential for breaking down silos, enabling seamless collaboration, and driving digital transformation. Below are best practices for achieving this synergy: 1. Establish a #UnifiedDataVision Capital One Capital One's DX revolved around adopting a unified data vision. By becoming a data-first company, it centralized customer and transactional data, enabling tailored offerings through #ML models. The initiative was backed by leadership and tightly integrated with their #digitalStrategy, leading to higher customer retention and improved personalization. 2. Design for #Interoperability Fannie Mae Fannie Mae built a #DataMesh architecture using Amazon Redshift data sharing. This decentralized approach allowed teams to access and share datasets without the bottlenecks of traditional centralized #DataWarehouse. It facilitated interoperability across tools and teams, enhancing operational decision-making. 3. Leverage Modern Architectures -Michelin Michelin's EFFIFUEL initiative utilized a #DataFabric to unify fleet data globally. The architecture provided real-time #insights into fuel efficiency and predictive maintenance. It enabled Michelin to offer #DataDriven advisory services to its customers while streamlining internal processes. -Fannie Mae, a leading provider of mortgage financing in the U.S., embarked on a journey to modernize its data architecture using a data mesh approach. The goal was to address challenges related to siloed data, enhance decision-making, and meet strict regulatory #compliance requirements. 4. Focus on #DataQuality and #DataGovernance Independence Health Group Independence Health Group implemented a robust data governance framework to safeguard the Protected Health Information (PHI) of 8.3 million insureds. By incorporating advanced encryption, access controls, and regular audits, they significantly minimized healthcare #DataBreach risks, ensuring compliance with HIPAA regulations. 5. Integrate Systems Through Middleware and #APIs Unilever integrated its global ERP, CRM, and supply chain systems using API gateways and middleware. This enabled real-time visibility into inventory levels and customer demand, reducing waste and improving operational efficiency. Middleware also ensured that legacy systems could coexist with new applications during the transition. 6. Utilize #Cloud Platforms for Scalability General Electric (GE) GE’s Predix platform unified #IoT and #IndustrialData into a scalable cloud ecosystem. This cloud-based system allowed real-time monitoring and analysis of industrial assets, enabling predictive maintenance and optimized operations in factories. (Continue in 1st comment) Details of the steps are available in the Premium Content Newsletter. Image Source: a16z #TransformPartner – Your #DigitalTransformation Consultancy
-
Abel Wang 赞了此动态Abel Wang 赞了此动态What separates the good from the great? 👇 The ability to execute clear, winning strategies. Here are 5 frameworks that top CEOs swear by: 1. VRIO Analysis ---> Evaluate your resources based on Value, Rarity, Imitability, and Organization. ---> Gain a sustained competitive advantage by scoring high on all four. 2. The Value Stick ---> Increase willingness to pay (WTP). ---> Reduce willingness to sell (WTS). ---> Maximize margins through customer delight and cost efficiency. 3. Blue Ocean Strategy ---> Stop competing in saturated markets. ---> Create uncontested market spaces by focusing on innovation, value creation, and cost reduction. 4. McKinsey’s Strategic Horizons ---> Horizon 1: Strengthen your core business. ---> Horizon 2: Nurture emerging opportunities. ---> Horizon 3: Innovate for long-term growth. 5. Hambrick & Fredrickson’s Strategy Diamond ---> Answer critical questions: ---> Where will we compete? ---> How will we win? ---> What’s our speed to market? These frameworks are the foundation for game-changing strategies. -- Want to create a winning strategy for your business? Follow Alan (AJ) Silber for more.
工作经历和教育背景
-
CCB Fintech
查看Abel的完整工作经历
查看他们的职位头衔、任职时间等。
或
资格认证
-
IBM Certified Senior ITA, also TOGAF
TOGAF, IBM
荣誉奖项
-
Best Paper of IBM Tech Summit
IBM
“Deep Thinking about SOA Practice ”,
语言能力
-
Chinese
母语或精通双语
-
English
中高级 (商务会话)
收到的推荐信
-
领英用户
2 位会员推荐了Abel
加入领英，即可查看其他相似会员
浏览更多动态
-
audax
𝗖𝗼-𝗘𝘅𝗶𝘀𝘁𝗲𝗻𝗰𝗲 𝗶𝘀𝗻’𝘁 𝗮 𝗰𝗼𝗺𝗽𝗿𝗼𝗺𝗶𝘀𝗲 — 𝗶𝘁 𝗮𝗹𝗹𝗼𝘄𝘀 𝗯𝗮𝗻𝗸𝘀 𝘁𝗼 𝗺𝗼𝗱𝗲𝗿𝗻𝗶𝘀𝗲 𝗳𝗮𝘀𝘁𝗲𝗿 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗱𝗶𝘀𝗿𝘂𝗽𝘁𝗶𝗻𝗴 𝗼𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀 𝗼𝗿 𝗰𝗼𝗺𝗽𝗿𝗼𝗺𝗶𝘀𝗶𝗻𝗴 𝘀𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆. 🚀 Our latest whitepaper, 𝙏𝙝𝙚 𝘾𝙤-𝙀𝙭𝙞𝙨𝙩𝙚𝙣𝙘𝙚 𝘽𝙡𝙪𝙚𝙥𝙧𝙞𝙣𝙩: 𝘽𝙧𝙞𝙙𝙜𝙞𝙣𝙜 𝙇𝙚𝙜𝙖𝙘𝙮 𝙖𝙣𝙙 𝙄𝙣𝙣𝙤𝙫𝙖𝙩𝙞𝙤𝙣 𝙞𝙣 𝘽𝙖𝙣𝙠𝙞𝙣𝙜, co-authored by Mathias Faure, Chief Product & Technology Officer at audax, and Balamurugan Jegatheesan, Chief Architect at FPT Software, explores how layering modern, cloud-native platforms on top of the legacy core enables faster innovation while keeping operations stable, secure, and compliant. ⚡ 𝗧𝗵𝗲 𝘄𝗵𝗶𝘁𝗲𝗽𝗮𝗽𝗲𝗿 𝘂𝗻𝗽𝗮𝗰𝗸𝘀: 👉 How banks can modernise incrementally while keeping core operations stable 👉 Why co-existence is becoming a strategic foundation for scalable digital transformation 👉 Practical frameworks for empowering domain-led teams and scaling innovation …and much more. For banks and financial institutions navigating legacy systems, this blueprint offers a clear, actionable path to long-term agility, sustainable innovation, and competitive advantage. ✨ 🔍 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗳𝘂𝗹𝗹 𝘄𝗵𝗶𝘁𝗲𝗽𝗮𝗽𝗲𝗿 𝗵𝗲𝗿𝗲: https://lnkd.in/d8j45PEv Fernn Lim Wenting Chen Man Ting Koh Gaurav GOUR Lisa Nguyen Minh Do #audax #Yeswecanlah #FPTSoftware #Coexistence #DigitalBanking #BankingInnovation #CoreBanking #LegacyModernization #Banking #API
1 条评论 -
FPT IS Careers
🚀 FPT Joins Qualcomm-Led Global Coalition to Accelerate AI-Native 6G Networks FPT has officially joined the global strategic coalition initiated by Qualcomm Technologies to advance AI-native 6G networks. The alliance unites leading tech and telecom companies to establish a clear roadmap: pre-commercial demonstrations in 2028 and commercialization starting 2029. This partnership strengthens FPT’s capabilities in intelligent transportation, autonomous UAVs, smart manufacturing, and advanced digital technologies. 💡 Chairman Truong Gia Binh: "This coalition will enhance FPT's ability to transform cutting-edge technologies into real-world value, driving sustainable impact for a smarter world." 🤝 This strategic move solidifies FPT's position in the global high-tech value chain and supports Vietnam's digital transformation agenda. #FPT #FPTIS #fptisscareers #fisnews