为了创造满足市场各种金融需求的创新产品，Piramal资本和住房金融于2022年12月15日在班加罗尔开设了Piramal创新实验室。这个3.6万平方英尺的创新中心将由该公司的首席技术官（CTO）Saurabh Mittal和Piramal商业智能部门负责人Markandey Upadhyay共同领导。
一、创新实验室将会提出哪些解决方案？你们会将哪些技术运用在这个解决方案上？
（一）采取“技术第一”的方法
作为一家公司，我们采取了“技术第一”的方法，这反映在整个组织的思维、功能和业务流程中。我们的理念是识别问题或机会，衡量它们，并构建技术解决方案来解决它们。有时我们会看到成功，有时我们需要一些迭代，但这是我们将追求的方法。因此，这个创新实验室的目的将是发现问题并为它们创造解决方案。
（二）创新实验室的目的将是发现问题并为它们创造解决方案
我们正在解决的一些关键问题包括为二、三线城市的客户创建保险解决方案。与一线城市的工薪阶层不同，小城市的大多数人可能是个体经营者，从事小生意。承销基础设施将使我们能够利用这些广泛客户群体的信息，将其输入某些预测模型，使我们能够大规模地做出信贷决策。
在这个行业中，还没有一个地方可以让你在几分钟内获得住房贷款。因此，我们正在努力缩短我们的住房贷款客户的周转时间，并为我们的无担保贷款客户提供即时决策和付款。
我们关注的另一个有趣的领域是银行对账单分析。我们收到各种格式的银行对账单，但在这个行业中没有一个单一的解决方案可以帮助获得客户的收入。
为了开发这些产品，我们将大量使用数据、人工智能和机器学习。通过新的创新中心，我们打算吸引产品管理、数据科学、用户体验和软件工程领域的熟练资源。该公司的目标是在23财年末建立一支由300多名技术专业人员组成的团队。
二、但对于一个相对较新的市场进入者来说，利用技术并迅速为公司创造竞争优势也很重要，你是怎么做到的？
（一）重新构想了这一过程，并将其转换为完全数字化的旅程
我们刚刚成立两年，在某些领域正赶上最好的公司。然而，与此同时，我们在很多领域都领先于其他国家。例如，我们有一个以纸张为基础的流程来签约直销代理，到实地为我们寻找业务。这个过程漫长而令人沮丧，花了七天时间才找到一个渠道合作伙伴。我们重新构想了这一过程，并将其转换为完全数字化的旅程。现在 DSAs注册平均只需12分钟。我听说这是行业首创。
（二）我们的API思想遥遥领先
然后我们有嵌入式金融合作伙伴。你可以把它们想象成各种各样的消费科技或金融科技公司，它们希望与我们合作提供贷款，寻求接触我们的客户。为了实现这一点，我们转向了API。后端的API栈使客户能够与贷方进行交互。嵌入式金融业务已经让我们的22个合作伙伴与领先的数字消费者和商家参与平台合作推出了超过24个项目。我们与合作伙伴上线的最快时间是四周，这也是行业第一的能力。我们的API思想遥遥领先。
我们的信贷经理会见潜在客户，并提出各种问题作为个人讨论的一部分。根据个人讨论的结果，信贷经理决定是否向客户提供贷款。我们在这个个人讨论的过程中加入了智慧。当信贷经理提出问题时，他会得到反馈，因为动态评分发生在后端。基于实时发生的评分，信用经理可以暂停、拒绝或批准客户。我认为这样的个人讨论工具还没有被其他玩家开发出来。
我认为我们很幸运，因为我们是一家年轻的公司，除了从DHFL收购中得到的东西外，我们没有很多遗留系统需要处理。当我们收购DHFL时，我们有一个已迁移到云上的就地数据中心。
三、这些解决方案和其他正在开发的解决方案将增加公司的收入，你们是如何通过技术提高利润的？
让我用集合的例子来说明这一点。收集可以完全以脱机方式进行。为了提高这一领域的效率，我们构建了一个名为Collection Central的智能应用程序。通过AI和ML模型，应用程序告诉我们，如果你给特定的客户发了一条消息，或者需要打电话或实地访问某个客户，该客户就会付钱。这确保了我们不会对每个客户都进行实地考察。这些解决方案在数据驱动的智能支持下，提高了效率。我很难说我们是否因为这些解决方案而收集了更多，但我可以肯定地说，因为它们，我们收集得更快，成本更低。
四、区块链为金融服务公司带来了希望，因为它可以带来更便宜和更快的交易，增强安全性和自动化合同，你怎么让Piramal的利益最大化？
Mittal:我们对区块链还没有积极的投资。区块链可以发挥重要作用的领域之一是财产注册商。鉴定房产文件的真伪，然后追溯它的遗产，从第一个买家一直到现在，这是一项艰巨的工作。建立一个行业中立的财产注册平台，由区块链支持，这让我们确保财产所有权是有效的，这是至关重要的，但像这样的用例将更多地是全行业的机会。因此，其中一些必须在数字贷款协会或其他论坛中进行，在那里你必须获得其他参与者的支持。
五、大多数金融服务公司在多个业务部门中都有数据竖井，您如何确保数据民主化以交付个性化的CX？
我们有一个单一的多产品平台，根据所使用的产品在内部分支成不同的流程。我们有一个单一的应用程序，所有的业务部门使用所有的产品，但它发挥不同，取决于他们开始的产品旅程。
我们确保我们所有的数据都生成并存储在一个地方，任何人都可以使用它。来自平台的每一个数据都流入一个数据仓库，该数据仓库为任何需要数据的人提供可访问性，无论是用于报告或可视化分析需求，还是用于在此基础上构建投影和机器学习模型。
我们已经规定，任何新的微服务或应用程序如果没有将所需的数据元素推入数据仓库，就不会投入生产。为了促进这一点，我们创建了一个“推送案例架构”，允许任何新的应用程序直接将数据推送到数据仓库，这使得开发人员和应用程序所有者非常容易做到这一点。
六、作为CTO，你面临的最大挑战是什么？
最大的挑战是招聘到我们想要的那种人才。大约一年半以前，我们公司连一个软件开发工程师都没有。我们从定义职位描述、角色、职责和吸引人才开始。我们在过去的一年半里取得了成功，但创新实验室现在将加速这一进程。
另一个巨大的挑战与我们在使用第三方系统时所面临的限制有关。我们有云原生的，并且设计的一切都考虑到了云。例如，从第一天开始，我们就使用无服务器计算和云管理数据库。除了按需供应、弹性和深度可观察性的好处之外，它还帮助我们专注于核心业务。
然而，第三方系统可能不是为云设计的，这给我们的战略和运营带来了瓶颈。我们一直在思考如何引入云原生思维来改进设置。
七、展望未来，这个行业最重要的商业和技术趋势是什么？
在贷款领域，账户聚合器是明年可能出现指数级增长的一个领域。这对于我们服务的大部分客户来说尤其重要，他们是信用基础的新手。我们没有这类客户的民事记录，需要他们可靠的银行对账单。因此，帐户聚合器以非常低的摩擦获得验证、验证和可靠的银行对账单是当前的需要。来自各个监管机构的推动已经在那里，在这方面，我们坚信账户聚合器将在明年的贷款领域成为一个主要的东西。
另一项将进入完全不同轨道的技术是机器学习。虽然所有贷款机构都基于内部和行业数据构建ML模型，但人们的思维正在向新的可能性开放。与ChatGPT, DALLE，以及我们周围的其他创新，有一组完全不同的机会出现，使用机器学习可以为客户和内部用户提供不可想象的体验。
原文：
To create innovative products that meet the various finance requirements of the market, Piramal Capital & Housing Finance opened the Piramal Innovation Lab in Bengaluru on Dec. 15, 2022. The 36,000-square-foot innovation hub will be led by the company’s CTO, Saurabh Mittal, and Markandey Upadhyay, head of business intelligence unit for Piramal.
CIO.com: What solutions will come up at the innovation lab and which technologies would you be leveraging for developing them?
Mittal: As a company, we have taken a ‘tech first’ approach, which reflects in the thinking, functions, and business processes across the organization. Our philosophy is to identify problems or opportunities, size them, and build technology solutions to address them. Sometimes we will see success, sometimes we will require a few iterations, but that’s the approach that we will pursue. The purpose of this innovation lab, therefore, will be to identify problems and create solutions for them.
Some of the key problems that we are working on include creating an underwriting solution for our customers in tier 2 and 3 cities. Unlike salaried people in tier 1 cities, most people in smaller cities may be self-employed and engaged in small businesses. An underwriting infrastructure will allow us to leverage information, available across these wide sets of customers, by feeding it into certain projection models that will enable us to take credit decisions at scale.
There still isn’t a place in the industry where you can get a home loan in minutes. So, we are working on reducing turn-around times for our home loan customers, and instant decisions and disbursements for our unsecured loan customers.
Another interesting area we are focusing on is that of bank statement analysis. We receive all kinds of bank statements in various formats but there isn’t a single solution in the industry that can help derive the income of the customer.
To develop these products, we will heavily use data, artificial intelligence, and machine learning. Through the new state-of the-art innovation centre, we intend to attract skilled resources in the areas of product management, data sciences, user experience, and software engineering. The company aims to build a team of more than 300 technology professionals by the end of FY23.
But for a relatively new entrant in the market, it is also important to leverage technology and quickly create a competitive differentiator for the company. How have you done that?
We are just about two years old and are catching up with the best in certain areas. However, at the same time, there are a whole lot of areas where we are ahead of others. For instance, we had a paper-based process to sign up DSAs [direct sales agents], go in the field and source business for us. The process, which was long and frustrating, took seven days to onboard a channel partner. We reimagined that process and converted it to a completely digital journey. Now DSAs get signed up in an average of 12 minutes. I’m told that’s an industry first.
Then we’ve got embedded finance partners. Think of them as various kinds of consumer tech or fintech companies, who want to give loans in partnership with us, seeking access to our customers. To enable this, we have turned to APIs. The API stack at the back end enables customers to interact with the lenders. The Embedded Finance business has allowed us to get 22 of our partners to launch over 24 programs in collaboration with leading digital consumers and merchant engagement platforms. The fastest that we’ve gone live with a partner has been about four weeks, which is also an industry-first capability. We are far ahead in our API thinking.
Our credit managers meet potential customers and ask various questions as part of a personal discussion.
Based upon the outcome of this personal discussion, the credit manager takes a decision whether the customer should be extended a loan or not. We have embedded intelligence into this process of personal discussion. As the credit manager asks questions, he gets feedback because of the dynamic scoring happening at the back end. Based on the scoring happening in real-time, the credit manager can pause and reject or approve a customer. I don’t think that such a personal discussion tool has been developed by any other player yet.
I think we got a bit lucky as being a young company we didn’t have a whole lot of legacy systems to deal with other than what we got from the DHFL acquisition. When we acquired DHFL, we had an on-prem data center that has been migrated to the cloud.
These solutions and the others in the pipeline will add to the company’s top line. How are you boosting the bottom line through technology?
Let me illustrate this with the example of collections. Collections could happen purely in an offline manner. To drive efficiencies in this area, we have built an intelligent app called Collection Central. Through AI and ML models, the app tells us that a particular customer will pay if you send a message to him or her or need to make a phone call or a field visit to a certain customer. This ensures we’re not making a field visit for every customer. Such solutions, supported by intelligence powered by the data, drive efficiencies. It’ll be hard for me to say whether we collect more because of such solutions but I can confidently say that we collect faster and with lesser cost because of them.
Blockchain holds promise for financial service companies as it can lead to cheaper and faster transactions, enhanced security, and automated contracts. How are you maximizing it for Piramal?
Mittal: We don’t have active investments in blockchain yet. One of the areas where blockchain can play a vital role is that of a property registrar. It’s hard work to identify the genuineness of property documents and then tracing its legacy all the way from the first buyer till now. Building an industry-neutral property registration platform, enabled by blockchain, that gives us assurance that the property title is valid is crucial but use cases like these would be more of industry-wide opportunities. Some of these, therefore, would have to be taken up within the Digital Lenders Association or other forums where you must garner support from other players.
Most financial services companies have data siloed in multiple business units. How do you ensure that data is democratized to deliver personalized CX?
We have a single multi-product platform that internally branches out into different flows depending on what product in being used. We have a single app that all business units use for all the products, but it plays out differently depending on which product they are starting the journey for.
We have ensured that all our data is generated and stored in a single place in a manner such that anybody can consume and use it. Every single piece of data from the platform flows into a data warehouse that provides accessibility of data to whoever needs it, either for a report or for visualization analytical needs or for building projection and machine learning models on top of that.
We have mandated that any new microservices or applications will not be put into production if they are not pushing the required data elements into the data warehouse. To facilitate this, we have created a ‘push case architecture’ that allows any new application to push data to the data warehouse directly, making it very easy for developers and application owners to do so.
As a CTO, what are some of the biggest challenges that you face?
Mittal: The biggest challenge has been hiring the kind of talent we would like to have. About a year and a half back, we didn’t have a single software development engineer in the company. We started by defining the job description, roles, responsibilities, and attracting talent. We had our success in the last year and a half, but the innovation lab will now accelerate it.
The other big challenge relates to constraints that we face while working with third-party systems. We have cloud native and have designed everything keeping cloud in mind. For instance, from day one, we use serverless computing and cloud-managed databases. Besides the benefits of on-demand provisioning, elasticity, and deep observability, it helps us to focus on the core business.
However, third-party systems may not have been designed for the cloud, which creates bottlenecks for our strategy and operations. We keep thinking how we can bring in the cloud native thinking there to improve the setup.
Going forward, what will be the top business and technology trends in this industry?
Mittal: In the lending world, account aggregator is one thing that is likely to see exponential growth next year. It is especially relevant for the large segment of customers we serve, who are new to credit base. We don’t have a civil record for such customers and need to have their reliable bank statements. So, account aggregator to get authenticated, verified, and reliable bank statements with very low friction is the need of the hour. The push from various regulators is already there, and on this account, we strongly believe that account aggregator will be a major thing next year in the lending world.
The other technology that will go to a different orbit altogether is machine learning. While all lenders build ML models based on internal and industry data, the mind is opening to newer possibilities. With ChatGPT, DALL.E, and other innovations around us, there is a completely different set of opportunities emerging and unthinkable experiences can be offered to customers and internal users using machine learning.
本文由数字化转型网翻译而成，翻译：数字化转型网默然。
免责声明: 本网站(http://www.szhzxw.cn/)内容主要来自原创、合作媒体供稿和第三方投稿，凡在本网站出现的信息，均仅供参考。本网站将尽力确保所提供信息的准确性及可靠性，但不保证有关资料的准确性及可靠性，读者在使用前请进一步核实，并对任何自主决定的行为负责。本网站对有关资料所引致的错误、不确或遗漏，概不负任何法律责任。
本网站刊载的所有内容(包括但不仅限文字、图片、LOGO、音频、视频、软件、程序等) 版权归原作者所有。任何单位或个人认为本网站中的内容可能涉嫌侵犯其知识产权或存在不实内容时，请及时通知本站，予以删除。
