In 2024, WNS rebranded and relaunched its analytics practice to include data, analytics, AI services, and consulting in a unified offering to enable decision intelligence. I spoke with Gautam Singh, the head of WNS Analytics, covering the opportunities and pitfalls of GenAI solutions, the team’s vertical and domain focus, and the company’s vision for productized services. We also discussed the requirements for core use cases to achieve quick ROI with data engineering and analytics.

WNS Analytics’ AI + HI approach

The current WNS Analytics practice is a result of organic growth and acquisitions, the latest being the 2022 addition of intelligence and analytics services pureplay The Smart Cube. Today, the practice comprises analytics experts, statisticians, data engineers and scientists, data and solution architects, business analysts, and consultants supporting ~250 clients, including Fortune 100 and Fortune 500, many shared with WNS BPS. WNS Analytics is organized into ten industries, the same as WNS.

Its core offerings include:

• Data processing and integration, data modernization, data engineering, cloud engineering, data quality and governance, and MDM
• Analytics, where it delivers data analysis, descriptive analytics, exploratory data analysis, predictive and prescriptive analytics
• AI and GenAI capabilities including ML, DL, NLP / NLG, computer vision, prompt engineering, AIOps, MLOps, LLMOps, and Responsible and Ethical AI.

It runs an AI lab to accelerate innovation for client organizations and a co-creation lab to enable advisory services. They are currently co-creating GenAI solutions with hyperscalers such as Microsoft and AWS.

WNS Analytics aims to combine proprietary AI-enabled assets such as accelerators, frameworks, and best practices with human intelligence (HI) by adding subject matter expertise across domains, data engineering, data science, AI and GenAI, analytics, and technology.

Turbocharging WNS with productized services

WNS Analytics looks to integrate AI, analytics, technology, and processes with domain understanding. It focuses on services that are best supported through productizing elements of the business processes but still need humans in the loop, domain understanding (even at the data engineering level), and process excellence. This way, the company wants to position itself as a more credible co-innovator.

Its target buyers of direct analytics services are Chief Analytics Officers, Chief Data Officers, and Chief Digital Officers. Some of the engagements include data engineering to modernize clients’ data ecosystem, building recommendation engines for sharper marketing campaigns, AI-driven competitive intelligence, and managing organizational risk using AI-driven fraud analytics. The remaining part (approximately half) of the business offers business decision-makers analytics built into BPS. Integrating advanced analytics into BPS is a significant GTM development where WNS will not separate and price analytics as a standalone. For example, in CX services, this model applies to agent enhancements.

One such deployment is for an online travel agency, which struggled with low customer experience due to long TAT caused by agents accessing lengthy knowledge articles and policies across multiple recommended results and requiring manual effort to decipher, contextualize, and summarize the information. WNS Analytics redesigned the agency’s knowledge management approach and deployed its GenAI-led Knowledge Assist platform. It introduced a graph database for efficient knowledge management across ~1k airline policy documents and enabled responses through GenAI, which simulated human conversations. It also built a user-friendly UI for the agent to pass queries and receive responses and enabled language translation. As a result, WNS Analytics delivered an 18-20% AHT reduction, lower latency and processing time, and increased the CSAT score between 3-5%.

One-to-many utilities library to power productized services

WNS Analytics has a utility library of industry-aligned AI/ML and technology components, tools and platforms across its ten verticals and domains. It has ~45 AI-driven products, platforms and boosters powering ~50 productized services. The utility library is available on the Unified Analytics Platform (UAP) architecture which is cloud-agnostic and supports structured, semi-structured and unstructured data. It consists of an ingestion layer which is API-driven and contains connectors for email, message, and gateway connectors to support various business data ecosystems. Domain specific analytical engines are enabled with AI/ML layer and metadata layer. The user consumption layer of the Unified Analytics Platform (UAP) encompasses a user interface which enables various domain-specific AI and analytics use cases, business workflows and visualization.

These components across the layers are interconnected and interdependent. They are configured together to form platforms and products such as SKENSE, Responsible AI solution, GenAI-led Knowledge Assist, and Unified Analytics Platform (UAP) for the insurance sector, for horizontal and domain-specific use cases. WNS Analytics works on the library to continuously build reusable components to enable multiple comprehensive use cases.

ROI on core use cases

Most organizations are still investing to get their data silos stitched together. WNS Analytics helps them start with core use cases to identify what data and analysis are needed and to build a data pond instead of an entire data lake. This quicker data engineering solution allows it to connect the different data ponds and technologies later. For example, WNS Analytics works with the business units in retail and CPG to tackle revenue growth management, customer experience management, customer and loyalty analytics, supply chain analytics, and digital analytics. Of these, enhancing revenues is at the top of retail executives’ minds.

For a multinational retail company, WNS Analytics applied its loyalty management best practices and customer 360 frameworks to hyperpersonalize loyalty campaigns. It designed new data management for improved data quality and triangulation of customer data. It built a single source of truth through a customer 360 framework, segmented customers for the loyalty program, and deployed AI-led personalized customer recommendations by analyzing association, taste, and look-alike profiles. Finally, it enabled GenAI personalized recommendation content. WNS Analytics mapped ~22m transactions to achieve a 10% increase in customer loyalty and a 400% increase in email click-through rate with faster personalization of the email content and increased speed to market for the content generation, from 15 days down to just ten minutes.

Part of the evolved GTM for WNS Analytics is absorbing some of the client risks through outcome-based commercial models. An example area is WNS’ recovery-as-a-service offering, which assists insurance companies in collecting auto claims unpaid from counterpart insurers. It analyzes the information, processes through image recognition photo evidence, verifies the claim veracity, and allocates an agent to collect. WNS Analytics only charges on the generated value for the insurer.

Addressing future market needs

WNS is evolving to meet the market needs for BPS undivided from the underlying technology and analytics and delivering better, faster, and cheaper outsourcing. It is strengthening the consulting bench with domain specialists with management consulting backgrounds and analytics understanding. These resources can start with the business case and are able to identify what hyperautomation, data, analytics and AI, combined with the process, are required to enable the business.

Another target skill set within the advisory talent is the ability to utilize GenAI, which combines lots of data at an unprecedented rate today. Enterprises are prioritizing multimodal LLMs and SLMs, paying attention to responsible AI, and looking for MLOps and LLMOps to operationalize and govern current and future GenAI solutions. In this environment, WNS Analytics looks for experts who will not use GenAI as a shotgun solution to solve any problems but as a pointed answer to subprocess-level challenges. WNS Analytics recognizes that GenAI is not the only answer for all business challenges.

At the same time, WNS is investing in infusing GenAI-specific accelerators and boosters in its tech stack; for example, by adding GenAI-based computer-generated summaries of patients’ medical documents in SKENSE. It has already identified ~100 GenAI use cases.

Clients are also looking to bring GenAI and BI together. WNS Analytics invests in accelerating BI creation and consumption through NLP, including BI virtual assistants. It is working on automated dynamic reporting, advanced data analysis and interpretation beyond the capabilities of traditional BI tools, and personalized and contextual data visualization. WNS Analytics is building upon its capabilities for visual storytelling to create compelling data stories using GPT and NLP libraries to combine visuals with narratives to influence decision-making.