How AI assistants can make sales teams more effective [Q&A]

Most business areas are starting to feel the impact of artificial intelligence, enabling data to be used more effectively to deliver intelligence and outcomes.

Sales, though, has always been about people and building relationships over time, so can AI offer value here too? Nikhil Cheerla, co-founder of sales AI assistant company Nooks, thinks it can, and we talked to him to find out how.

BN: Tell us about the specific challenge that sales teams have that AI can help solve?

NC: Sales teams operate at the intersection of structured data and human interaction, making their workflows uniquely complex and highly variable. This complexity creates two core challenges. First, sales reps are overwhelmed by fragmented data spread across CRMs, emails, and external sources like LinkedIn or industry databases. Synthesizing this into actionable insights often consumes valuable time. Second, repetitive tasks such as dialing, following up, and manually logging activity steal focus from what reps do best -- engaging with prospects.

AI assistants are designed to address these inefficiencies by transforming raw data into actionable insights and automating repetitive tasks. For example, a sales rep preparing for a call typically spends time pulling a prospect's history from the CRM, doing pre-call research, and dialing multiple times, often reaching voicemail. AI can streamline this by integrating data across platforms, providing a concise account summary in real time, and handling multi-line dialing with automatic voicemail drops.

When the call connects, the assistant goes further, delivering tailored recommendations based on the prospect's profile, past interactions, and buying signals. This enables reps to have more meaningful conversations, driving better outcomes and pipeline growth. AI assistants act as intelligent intermediaries, augmenting the science of sales while allowing reps to focus on the art -- building trust and relationships.

BN: Why AI assistants, not agents?

NC: AI assistants are reactive -- they respond to user commands or queries and require direct interaction to perform tasks. On the other hand, AI agents are proactive and autonomous. They make decisions and take actions independently, without constant user input. This distinction is important.

AI assistants are designed to augment human workflows. They provide insights, automate specific tasks, and respond to human direction while leaving critical decision-making to the sales rep. For example, an assistant might surface the most relevant prospect information and draft a personalized email, but the rep decides how to frame the conversation. By contrast, agents operate autonomously, making independent decisions and taking actions without human oversight. This autonomy can be risky in sales, where understanding tone, intent, and nuance is critical.

One of the biggest risks with autonomous agents is their susceptibility to hallucinations -- generating plausible but inaccurate outputs. This makes them unsuitable for roles like sales, where trust is paramount. AI assistants, on the other hand, act as collaborators, ensuring reps stay in control of decisions while benefiting from the computational efficiency of AI. By prioritizing assistants over agents, Nooks ensures that sales workflows remain flexible, reliable, and human-led.

BN: What are the technical challenges in working with AI assistants?

NC: Building AI assistants involves addressing several key areas. One of the most significant is data aggregation and normalization. For sales, for example, data comes from diverse sources -- CRM records, web scrapes, email logs -- each with its own format and context. AI assistants have to be able to integrate and standardize this data into a cohesive, usable format.

Another challenge is making them perform in real-time. Sales reps need insights during live calls or meetings, which demands low-latency responses. Achieving this requires sophisticated caching, indexing, and inference pipelines to ensure reliability and speed. Contextual understanding is also critical. Assistants must parse structured data like CRM entries and unstructured data like email threads to extract relevant insights, requiring advanced natural language processing and semantic analysis capabilities.

One of the more nuanced technical hurdles is hallucination mitigation. Large language models can occasionally generate inaccurate or misleading outputs, which is unacceptable in high-stakes interactions like sales. AI assistants need to incorporate guardrails, such as confidence scoring and data lineage tracking, to ensure outputs are reliable and verifiable.

Let's not forget scaling so the assistants grow with a business. As data volumes increase and workflows evolve, assistants must remain responsive and accurate.

BN: Do you see AI as complementary to humans or do you see the technology replacing them?

NC: Unlike deterministic tasks such as support ticket resolution or code generation, roles involving human interaction require adaptability, emotional intelligence, and context-aware decision-making.

In many AI workflows, the 'consumer' actively seeks the outcome, like a support solution or a generated piece of code. However, in less structured, interaction-driven scenarios, participants may not have clear expectations, and interactions are often subject to variables such as tone, cultural context, and evolving intentions. This creates an environment where the rigid frameworks of traditional AI systems struggle to perform effectively.

As I mentioned previously, AI systems are also prone to hallucinations -- generating plausible but incorrect outputs -- and often lack the capability to interpret subtle cues like tone, intent, and contextual nuances. These limitations make it risky to rely solely on AI for roles or tasks where authenticity and trust are paramount.

For AI to excel in these complex scenarios, it must operate as a collaborative tool, augmenting human capabilities rather than seeking to replace them. This requires addressing its reliability issues and building systems that enhance human creativity, empathy, and decision-making while maintaining transparency and accuracy in its outputs.

BN: If you project 10 years out, what does the relationship between humans and AI look like?

NC: I think the relationship between humans and AI will likely evolve into a seamless partnership where AI augments human capabilities across industries. Instead of replacing human roles, AI will take on routine, repetitive tasks -- such as data entry, research, and administrative workflows -- allowing people to focus on higher-value, creative, and strategic activities.

We envision leaner, smarter teams powered by AI systems that actively execute parts of their workflows -- analyzing data, providing personalized recommendations, and delivering real-time insights that elevate decision-making. AI will become indispensable in roles where efficiency and precision matter, functioning as an intelligent collaborator that enhances human output and drives better outcomes.

For example, in collaborative environments, AI could handle the background research for a complex project, leaving humans free to engage in problem-solving and relationship-building. Similarly, in technical fields, AI might synthesize vast amounts of data into actionable insights, enabling experts to make informed decisions faster.

AI is about empowering individuals to focus on what humans do best -- creativity, empathy, and complex decision-making. With AI handling the mechanics, humans can prioritize tasks that require judgment, innovation, and connection. This dynamic partnership will redefine roles across industries, setting a new standard for efficiency, growth, and meaningful work.

Image credit: [email protected]/depositphotos.com