Our Journey with Generative AI: Why We Built a Domain-Specific Research Bot

As a product lead with extensive experience in creating digital solutions in market and competitive intelligence, I’ve witnessed how the volume of insights has increased multifold in the last few years. My clients are primarily professional services firms, and this sheer volume of information available across various platforms makes it difficult for businesses to act swiftly and effectively.

Business leaders require prompt, precise, and targeted insights to inform critical actions and drive strategic decisions. However, their questions are usually unstructured and never seem to align with their pre-defined taxonomies. Not many existing tools handle this well.

That’s why I’m ecstatic to introduce Research Bot, Evalueserve’s groundbreaking generative AI chatbot. The conversational chatbot has been a long time coming for Evalueserve, as it aligns perfectly with our mission to enhance and accelerate client decision-making.

The chatbot leverages large language models (LLMs) and is trained on proprietary domain-specific data from Evalueserve platforms such as Insightsfirst, allowing it to give human-like responses that are professional, concise, and relevant to the specific task at hand.

We’re thrilled by the positive client reaction to Research Bot. Here’s a quick preview of how it works https://youtu.be/0DtVh-w_fKY

This blog delves deeper into why we developed Research Bot and how it helps clients hone in on actionable insights.

 

The First Live Demo

When our team first introduced Research Bot to a client, we were all a bit apprehensive. The client was a leading management consulting firm, and we were demoing Research Bot to their competitive intelligence team, the same team that uses our Insightsfirst Competitive Intelligence platform. Of course, the expectations were high!

With the widespread adoption of ChatGPT and money pouring into new generative AI products, we understood that we had one chance to prove that our chatbot was worth their time.

That’s why we made the bold decision to ditch the pre-scripted demo and let the client come up with questions on the spot. It was a gamble we felt was necessary to take.

Thankfully, it paid off.

The client was able to truly experience and test Research Bot’s domain expertise on their terms. It answered each question accurately and with nuance, outperforming all our expectations.

Here are screenshots from the actual conversation.

As you can see, the chatbot demonstrated deep domain expertise in specific topics and niches that the user was particularly interested in. It immediately summarized, collated, and shared the latest insights about key competitors, and then it allowed the client to drill into the sources. This process took a fraction of the time for an analyst to receive the request, conduct the research, and send it back to the client. 

The competitive intelligence director immediately saw that Research Bot would be a game-changer. Its directional insights would give analysts a head start in understanding key topics before diving into the next level of detail. The chatbot would significantly increase the speed of getting relevant insights and reduce the risk of missed opportunities.

Our Journey in AI

Before diving into the specific features and benefits of Research Bot, we wanted to step back and reflect on how we got here.

Research Bot didn’t just happen overnight.

While the actual development of the chatbot took only a few weeks, identifying and implementing a scalable, domain-specific solution for high-impact client use cases took months of work and substantial investment.

There are three key reasons why Evalueserve felt we were well-positioned to operationalize generative AI for enterprise use cases quickly:

  1. Evalueserve sits on a proprietary set of expert-curated data. This data is clean. While it covers a broad range of topics and industries from public and private sources, it caters to the use cases of strategy, knowledge, go-to-market, and product teams seeking a competitive edge.

    The angle is consistent, and the types of analysis and recommendations are designed for common enterprise decision-making scenarios: Does our strategy take advantage of trends in the industry? Are we addressing what competitors are talking about? Who should we be targeting? What are they currently interested in, and how can we reach them? Etc.

  2. We have thousands of domain experts and years of experience. Evalueserve teams provide market and competitive intelligence to Big 4, Tier 1 Management Consulting, leading Asset Management, and Fortune 1000 companies. We understand the most common types of insights requests, when they need them, and what they’re used for.
  3. Our solution architecture is designed for technology and AI. We’ve been infusing AI and automation into our solutions for years. At the core of this strategy is our AIRA platform, a hub of reusable, configurable microservices that can be used to build everything from internal accelerators to partner plugins and low-code platforms. This infrastructure is what enables us to seamlessly integrate the latest technologies like GPT into our products and services.

Without the data, expertise, and a robust technology strategy, we wouldn’t have been able to dive into the generative AI race so quickly.

Real-World Applications and Benefits

One of the most popular use cases for Research Bot is pulling targeted insights from our market and competitive intelligence platform, Insightsfirst. For example, a prominent international consulting firm is leveraging Research Bot to stay ahead of the competition’s media moves.

Keeping up with rapid news cycles was becoming increasingly difficult and time-consuming, especially across social media channels. To address this, they started using Insightsfirst to automate the process of monitoring, categorizing, and summarizing competitors’ posts. This news feed was already highly curated to their interest. Now with an AI chatbot, analysts can immediately chat with Research Bot to answer specific questions.

Across competitive intelligence teams, Research Bot can:

  • Provide the latest, most relevant insights tailored to your specific questions whenever you need them
  • Offer easy access to sources for each insight
  • Enable filters to focus on critical insights and narrow the scope of the conversation

Conducting research with a domain-specific chatbot allows you to cut through the noise and retrieve directional insights quickly. By accelerating the time to insights, companies can scale their ability to make data-driven decisions and equip themselves to gain a competitive edge.

Besides competitive intelligence, we’re witnessing numerous potential use cases for Research Bot. For instance, the chatbot can sift through thousands of previous RFPs your company has participated in, helping you include crucial details in your current submission. 

Domain-specific AI chatbots can also support processes like credit lending, where analysts must extract and verify information from various sources to assess risks and approve borrowers. Research Bot could quickly help uncover specific information that is difficult to find.

It’s safe to say, this is an exciting time for product leaders across industries. The capabilities of LLMs have far surpassed what we expected just one year ago.

I look forward to exploring the wide range of generative AI use cases across our diverse client base and am glad to be working at a place like Evalueserve, which is highly committed to innovation in applied AI. One of our goals as a company is to truly reimagine and transform how clients leverage and consume data. We believe we’ve reached a pivotal moment where technology can be harnessed to multiply the impact of insights and analytics.

We’d love to hear from you if you’re using generative AI in your analytics and insights programs. Just fill out our “Contact Us” form here https://www.evalueserve.com/speak-to-expert/. Make sure to include a brief description of your use case so we know why you’re reaching out.

Also, don’t hesitate to contact us if you want to stay abreast of our roadmap for Research Bot or other Generative AI features. We would be happy to have an exploratory discussion.

Satyajit Saha
Vice President and Product Lead, Professional Services Posts

Latest Posts