"We’re moving from keyword matches to conversations with data"

How has enterprise search evolved in India with the rise of generative AI? 

Enterprise search is changing thanks to generative AI as keyword-based and rigid approaches are being replaced with smarter and contextualized search experiences. Traditional search, which is commonly a "one-size-fits-all" process, often did not capture user intent, especially when using vague terms or faulty spelling in keyword searches.

Improved methods of natural language processing, machine learning applications like semantic search or Retrieval-Augmented Generation (RAG), give development teams the capability to enable search engines that acknowledge meaning instead of only word match searches.

By employing large language models along with access to proprietary data in real-time, enterprise search is more accurate with better and more relevant results. Interestingly, 44% of global IT leaders have said, in accordance with Elastic, that they expect GenAI to lead to increased productivity by saving up to two workdays per week.

What sets Elastic’s Search AI apart from traditional search approaches, especially in using LLMs? 

Traditional search engines may quickly retrieve information, but they cannot recognize or respond to user intent. In comparison, AI can analyze complex patterns, but it cannot locate specific information. Elastic’s Search AI platform combines strong search with generative AI to create actionable insights from abandoned, unstructured data that improve decision making, security and customer experience.

Elastic is open and LLM agnostic meaning it can effortlessly work with existing GenAI tools like LangChain and we give organisations the ability to use any large language model, either proprietary or open source.

Developers get access to Elastic’s production-grade vector database, this enables developers to understand structured and unstructured data in new and deeper ways which creates strong leverage for GenAI use cases ranging from smarter recommendation engines, to hyper-personalized customer engagement.

What challenges do Indian firms face when adding LLMs to hybrid or multi-cloud setups? 

Bringing large language models (LLMs) into hybrid or multi-cloud solutions has many benefits for enterprise search, but there are challenges. Whether operationally with the different APIs and management tools across cloud providers, as simple access or high-quality data is what both key factors of success are built upon, dissolved between systems.

Organisations must also adhere to complicated regulatory compliance practices, such as India's Digital Personal Data Protection Act (DPDPA) 2023, which has strict consent requirements, prevents onward transfers between countries with exception; therefore, where compliant data is relevant to their LLM in regulated countries, it is impactful.

How do you ensure accuracy and minimise bias in AI-powered search results? 

Elastic's multi-layered approach to maintaining the accuracy and fairness of AI Enhanced Search integrates advanced technology, ethical governance, and strong security. Elastic's Search AI product merges traditional keyword search with both semantic understanding and hybrid ranking to minimize bias and maximize content relevance.

By embedding generative AI outputs with real world data using Retrieval-Augmented Generation (RAG) techniques, Elastic reduces hallucinations and misinformation. Additionally, Elastic aligns its practices to NIST AI Risk Management Frameworks and deploys built-in protections against attacks from threat vectors such as prompt injection and data leaks. Data Quality Dashboard tools also ensure that users have access to clean data which enables trustworthy AI results.

How does your platform address data privacy and compliance in regulated sectors like BFSI and healthcare?

We adopt a "secure by design" approach, ensuring it deploys and uses its own solutions—such as Elastic Security—in the first place to proactively detect and eliminate threats. This commitment to trust and compliance is particularly important to regulated industries like BFSI and healthcare.

Elastic is designed with customer compliance in mind and helps customers comply with local and global regulations such as GDPR, India's DPDP Act, RBI and SEBI guidelines as well as USA-based regulations like HIPAA.

Elastic provides security features including role-based access, audit logs, real-time monitoring of threat levels, and anonymisation and pseudonymisation to ensure data privacy, through ISO-certified security practices and procedures (ISO 27001, 27017, 27018), as well as a strong Governance policy to protect customer data from leaking or misuse.

What key innovations are you bringing to Search AI for developers in India? 

Elastic is allowing developers to innovate at a higher velocity and be more productive with its unified Search AI platform. The platform streamlines important tasks like log analysis, anomaly detection, and data correlation across on-prem and cloud environments, reducing manual work and allowing developers to focus on performance, security, and user experience.

Solutions such as Elastic Observability and Attack Discovery ensure issues are identified proactively and important threats are prioritised. With India rapidly transforming through digital advances from initiatives such as Digital India, Elastic's developer-first philosophy opens the door for developers to unlock unique and immersive GenAI-driven, secure, and scalable digital experiences.

Where do you see the biggest GenAI–search opportunities for Indian IT service providers in the next few years? 

GenAI-driven search is emerging as a powerful capability in BFSI, healthcare and the public sector, with significant implications as it becomes available for semantic search within documents as well as summarisation and conversational access to complex datasets. Technologies that may be incorporated into GenAI-driven search such as faceted search functionality in Elasticsearch provide an enhanced user experience that empowers searchers and enables AI-generated query classifications to drill deeper into insights.

The power of this is a feedback loop where better search leads to better AI outcomes, meanwhile, improved AI models will also result in better search, and will be used more frequently as it has become more responsive.

Tools like the Elastic AI Assistant as part of the Elastic AI ecosystem (for RAG applications, etc.) position Indian IT service providers to lead the next evolution in intelligent, scalable and inclusive search experiences.

Can you share notable success stories in AI-driven search adoption in India? 

AI-driven search has played a pivotal role in enhancing productivity, scalability and creating alternative revenue streams for Indian organisations, with one being Apna. Elastic’s Search AI solutions have helped Apna to become one of India’s fastest startups to gain unicorn status in the country, where it achieved a valuation of $1.1 billion.

As India’s largest job and professional networking platform, it has successfully matched organisations with prospective employees. Apna also provides live tips on job application, the interview process and assessments that save significant time for employers by screening applicants.

With Search AI, the platform is able to make sense of unconventional phrasing and better understand the user intent of searches conducted by candidates and employers. The increased accuracy of search results has led to Apna seeing a 15-20% increase in job applications and an 80% increase in profiles downloaded by employers on the platform, resulting in new revenue streams for the company.