MarTech Interview with Adit Abhyankar, CEO and co-founder of Breakthrough

Adit, please share your professional background and the journey that led you to co-found Breakthrough and eventually become the CEO.
I have loved tech since the dotcom boom days in San Francisco. Breakthrough is my seventh startup – other than a 5 year stint at Google, my entire career has been building B2B startups. From the moment I started hacking AI tools, I knew this was going to be huge and I realized I had to be part of this wave.

I co-founded Breakthrough with my CTO, Kevin, in early 2024. Our journey began with a fundamental question: Can AI truly sell? Instead of focusing on automating sales processes, we zeroed in on the heart of selling—the pitch. A great pitch isn’t just about delivering information; it requires a deep understanding of your value proposition, thorough prospect research, and authentic communication that builds trust.

When assessing the market, we identified two major challenges that lead to increased spam and slow moving or stalled deals. First, most AI-driven sales tools emphasize automation over actual sales communication. Second, these tools prioritize scale and efficiency, relying on surface-level research and generic messaging rather than refining persuasive, high-quality interactions. Our platform is addressing these by continuously learning from customer interactions to refine how businesses communicate their value. Rather than just automating sales, Breakthrough creates the most precise and effective messaging model at an enterprise level.

How do you see AI transforming sales and marketing strategies over the upcoming 5 to 10 years when it comes to personalized customer engagement?
Success in sales depends on identifying the right people to sell to and convincing them to trust you with their business. The key to using AI is to not just automate the research and thinking needed, but to also rapidly and accurately produce the right messaging to tell each story fantastically well, while continually learning from every engagement. Having spoken to hundreds of businesses, we know there is a wide-open opportunity to help improve the impact of every sales meeting, by improving the quality of business storytelling at scale.

By embedding AI in the entire sales meeting workflow – preparation, strategy, messaging, and automated follow-ups, all powered by the most accurate messaging model, sales leaders can unlock AI’s true potential for their teams.

How can companies use AI tools like Breakthrough to balance automation with maintaining authentic, human-centered messaging in customer relationships?
For 82% of global B2B marketing decision-makers, buyers expect personalized experiences but delivering them effectively requires significant time and effort. As a result, many sales teams are turning to generative AI (GenAI) to support their efforts. However, often the results fall short – requiring so much manual refinement that sales cycles are actually prolonged. Tools like ChatGPT will always do what you ask; they will never ask if you should be doing those things at all. They will never tell you they do not have enough of the right information to give you a good answer, assuming you know exactly what you want to ask and ask it the right way.

Tools like Breakthrough take this a step further, developing a deeper understanding of your business and strategy as well as that of your audience to help you align your pitch in a way that builds trust. We continually learn from every customer interaction to see what exact words they use to describe pain points, raise objections, etc. and this feeds our messaging models so sales executives can maintain authenticity and enhance, not erode trust, when using AI tools.

What steps can organizations take to make AI skills accessible beyond engineering teams and enable cross-functional collaboration
At least in our domain, sales teams are using AI tools, with or without the blessing of their organizations. Many sales people we talk to are taking AI courses on things like prompt engineering so that they learn how to get the most of these amazing capabilities. On some level, people are going to upskill anyway, as it helps them become way more productive and save time.

However, organizations have a role to play by actively encouraging their teams to benefit, instead of being overly risk averse or reluctant to start proof of concepts, even their senior leaders to roll up their sleeves and experience specialized AI tools first hand. Finally, looking at L&D (learning and development) opportunities that will provide exposure to a broad range of technology instead of needing to commit too much time to any one tool or provider.

The pace of advancement is so rapid that it can sometimes be bewildering and organizations may not know where to start. For me, it doesn’t matter where you start as long as you start getting your hands dirty and building confidence in navigating the technology. AI models are inherently collaborative because they are trained in all business domains, out of the box; it’s organizations that typically have the silos, and AI provides a very convenient enabler to break these down.

How can AI-driven sales and marketing tools enhance revenue growth for organizations , and what long-term value do they offer?
One of the biggest opportunities for AI in sales is, ironically, the ability to truly listen as many sales teams will ‘wing’ their meetings, instead of truly preparing for them. In the meeting, they will typically talk too much and assume what they are saying is resonating, leaving no room for their audience to imagine and ask questions, or allowing the scarce time available with their prospect to be diverted into the wrong conversation. This results in stalled sales pipelines or meetings that go nowhere.

AI tools today can truly “listen” by doing research but also by ingesting actual customer interactions and then automatically drawing insights that can uncover opportunities from the prospect’s point of view. New reasoning models are extremely capable at piecing different information together to draw the right conclusions and causal inferences; new writing models are very good at translating these insights into a well articulated pitch that will resonate. Furthermore, AI models can now balance the flexibility needed to truly align your pitch to your target stakeholder while still staying on message from a marketing and product point of view. The use of AI tools, done well, can be used to ensure sales people don’t oversell or undersell and to help them be positioned as knowledgeable and trustworthy.

Sales is a very natural domain to deploy AI models – it’s all about language and communication, which is really what LLMs are all about. I believe in the long term they will be deeply integrated into virtually all sales workflows, as selling is often one of the most expensive and unreliable parts of many organizations and AI will have a huge impact on both.

As AI personalizes sales and marketing, what ethical considerations should companies keep in mind regarding customer data?
When it comes to AI and personalisation, companies must balance innovation with ethical responsibility. With global AI governance diverging—some regions prioritizing innovation exemplified by models like DeepSeek and Manus AI, others emphasizing regulation—businesses must proactively establish their own trust frameworks. Ethical AI isn’t just about compliance; it’s about fostering long-term confidence and trust with a diverse set of stakeholders.

Transparency is key in AI development and deployment to ensure trust – customers need clear understanding and control over how their data is used and feel like they are controlling AI, not the other way around. Companies should prioritize data security, informed consent, and algorithmic fairness to prevent bias and misuse with AI-driven personalization, ensuring that outputs remain ‘explainable’ and that data is handled responsibly. Long-term success depends on balancing AI’s potential with ethical safeguards. Companies that embed trust, accountability, and fairness into AI-driven sales and marketing will not only comply with evolving regulations but also build stronger customer relationships.

How can AI help businesses create more relevant, nuanced, and compelling sales narratives that build trust with potential clients?
Trust is key to establishing a selling relationship. Building trust for a sales person involves showing that you are credible personally, and as an organization. AI can help a sales person do that by automating all the research and synthesis needed to understand your client, so you can align to their needs. AI can automatically understand your own products, positioning, differentiation and competitive landscape. And with all this context, it can guide a sales person towards the best strategy for each customer interaction. We sometimes refer to this as ‘Waze for sales’ – helping a sales rep define the strategy of how best to accomplish a sales goal, and then providing step by step guidance to help them execute against that strategy.

What strategies should businesses use to fund and support high-risk AI ventures with long-term potential?
Businesses should first understand the sources of potential risk and whether they truly are high risk or not. I don’t think businesses should actually undertake high risk with AI – but there are many ways to mitigate risk. Proof of concepts that don’t require deep system level integration is one way to do that; if you understand the value first, then you can work with a vendor to mitigate truly unacceptable risk.

With AI, in many cases, perceived risk is far greater than actual risk. The major model providers like OpenAI do not use enterprise customer data for training and do not share data across customers. This is not that dissimilar to working with a cloud provider like Google or Amazon. Most startups are built on top of the security and infrastructure of these larger providers and therefore ‘inherit’ all the data protection and data security policies.

How do platforms like Deep Search and DeepSeek impact data-driven sales and marketing, and how can businesses stay ahead of these advancements?
Platforms like Deep Research and DeepSeek are reshaping data-driven sales and marketing by making AI-powered insights more accessible, efficient, and cost-effective. These tools lower entry barriers, allowing businesses to leverage AI without massive infrastructure investments. Their open-source approach democratizes AI, making advanced data analysis and personalization capabilities available to a wider range of companies. However, this shift also intensifies competition, forcing businesses to rethink their AI strategies, and remain competitive while upholding their values.

To stay ahead, companies must prioritize proprietary data assets and AI-driven decision-making frameworks. While low-cost AI models offer opportunities and flexibility, companies must also assess the trade-offs as they require investment in infrastructure and talent. The key is striking the right balance between cost, control, and scalability.

AI in sales and marketing will no longer be limited to large enterprises with deep pockets. With the cost of AI models dropping significantly, value is shifting toward unique business logic, data and execution capabilities. In other words, how good your product is , how differentiated it is, how clear your strategy is, and how well you are meeting your customer needs. AI will help companies that have great products but poor selling teams, catch up and beat companies that have poor products and great selling teams.

What advice would you give to businesses looking to adopt AI tools in their sales and marketing strategies for the first time?
I would say focus on quality, not scale. Pick the AI platform that best models your business and strategy, not generic superficial AI platforms. When working with a vendor, first focus on understanding value and then on mitigating risk. How you construct and define an initial engagement can affect this greatly. I would also advise to not be prematurely worried about scaling and automation; understanding exactly how you get AI to improve business outcomes should be the priority.

One thing businesses should not do is simply go with the incumbent or generic AI tools like ChatGPT, assuming they will have the best product or be the most trustworthy vendor. The reason is that there are lots of AI-first startups bringing fresh thinking and innovative approaches to the space, and businesses should take a look at what best fits their needs.