AI is becoming central to how modern enterprises function. Tools like Salesforce, HubSpot, and Snowflake each have distinct structures, APIs, and business logic. In a freewheeling chat with CXO Media & APAC Media, Jatin Sandilya, Head of AI, Ampersand, elaborates on the challenge to enable AI agents to understand and act across all these systems without the need for complex reengineering.
How would you describe your vision for the role of AI in modern enterprise technology?
AI is becoming central to how modern enterprises function. At Ampersand, we see it as the layer that connects intent to execution. Our focus is on building intelligent agents that understand business goals, work across existing systems, and take action in real time.
These agents are designed to support teams in doing more with greater clarity and less effort. Whether it involves following up with leads, resolving customer issues, or coordinating internal processes, AI should work within current workflows and enhance them meaningfully. The goal is to make AI genuinely useful in everyday operations, not just a futuristic idea.
What are some of the key challenges you faced when designing AI systems that integrate seamlessly across diverse platforms?
Every platform in the enterprise stack operates differently. Tools like Salesforce, HubSpot, and Snowflake each have distinct structures, APIs, and business logic. One of the biggest challenges was enabling AI agents to understand and act across all these systems without the need for complex reengineering.
To solve this, we created the Model Context Protocol. It allows agents to interpret a business objective and translate it into structured actions that align with how each platform works. This gives teams the flexibility to deploy AI across systems without slowing down or rewriting workflows.
How is Ampersand approaching the shift from traditional automation to more intelligent, adaptive systems?
We are building agents that manage complete workflows rather than just automate individual tasks. Traditional automation usually follows rigid rules and breaks down when context changes. Our approach focuses on adaptive systems that can respond to real-time inputs and adjust accordingly.
For example, an agent managing sales follow-ups can prioritise leads based on engagement, modify outreach content, reschedule missed connections, and escalate when needed. These agents are designed to think through outcomes and learn continuously, which allows them to operate with greater independence and effectiveness.
What steps do you take to ensure AI solutions are not only innovative but also practical and scalable?
Innovation has to translate into results. That is why we start with the user problem and design solutions that can be implemented quickly and scaled easily. One of the most effective tools we have created is amp.yaml. It lets teams define agent behaviour in plain language, so they can make updates or launch new workflows without needing specialised skills.
In addition to ease of use, our infrastructure is modular and robust. Businesses can start small and expand with confidence, knowing that the underlying system can grow with them without needing to be replaced or rebuilt.
How do you see AI transforming customer engagement in SaaS platforms in the next few years?
Customer engagement is becoming more intuitive and more immediate. Users expect quicker answers and more relevant interactions. AI will make that possible by offering context-aware support, anticipating needs, and enabling real-time conversations.
At Ampersand, we help SaaS platforms introduce intelligent agents that handle onboarding, support, and upsell opportunities. These agents draw from usage patterns, behavioural data, and contextual signals to deliver personalised interactions. This creates a more responsive and effective experience, which directly impacts retention and growth.
Revert was centred around deep-tech innovation. How do you continue to bring that innovation-focused approach to your current role at Ampersand?
Revert focused on building a distributed infrastructure that handled real-time orchestration and data synchronisation. As co-founder and CTO, I worked on systems that needed to be performant, scalable, and dependable under pressure. That experience taught me how to architect for reliability and flexibility at the same time.
At Ampersand, I bring that mindset into how we design our agent systems. We still solve technically complex problems, but we also focus on making our tools usable by product and operations teams, not just engineers. Innovation is only meaningful if it leads to adoption and real value, and that continues to guide how we build.
What is your approach to balancing proprietary development with open-source collaboration in the AI space?
We take a hybrid approach. Open-source helps us move quickly, encourages community feedback, and gives teams more flexibility. That is why we make components like SDKs and connectors open for developers to build with and customise.
Our core orchestration and memory systems remain proprietary. These are critical to ensuring performance, reliability, and enterprise-grade support. By maintaining this balance, we can stay responsive to the ecosystem while protecting the stability our customers expect.
How does your GTM strategy ensure alignment between product innovation and real-world business adoption?
Product and go-to-market work together from the earliest stages. Every feature we launch is tested in real-world environments, and feedback from users is factored directly into our development process. This ensures that we are solving problems that matter, not just building what is possible.
We also invest heavily in enablement through documentation, templates, onboarding support, and live use cases. This helps teams adopt and expand quickly, creating a smoother path from innovation to business value.
In a competitive AI landscape, what sets Ampersand apart from other solution providers in terms of technology, integration, or value delivery?
Our agents are designed to deliver outcomes with autonomy and integrate deeply into existing tools. Many AI solutions automate small tasks or surface insights, but still rely heavily on manual coordination. Ampersand agents go further. They understand business intent, coordinate across systems, and execute full workflows independently.
That level of intelligence, combined with seamless integration and ease of configuration, allows us to deliver results faster and at a larger scale. This is where we bring the most value to our customers.
