Sree Unnikrishnan

Technology Capability

What is actually possible — and what is not (yet).

What the system can do has changed dramatically. From rule-based systems to probabilistic AI, capability is no longer the constraint — clarity is.

Most product failures come from misreading capability boundaries.

Human Behavior

What users actually do — not what we expect them to do.

People don't behave like flows or personas. They hesitate, adapt, mistrust, explore, and learn. Most products fail because they model logic, not behavior.

Adoption, trust, and retention are behavior problems, not feature problems.

Product Systems

How decisions scale over time.

Features don't scale. Systems do. The real work is designing feedback loops, participation models, and evolution paths — not screens.

The real work is defining systems that hold under growth, variation, and misuse.

Discovery

My earliest work was on machine learning driven discovery systems before "AI" became a product category.

At PurpleYogi and Stratify, the problem wasn't search. It was making information explorable. We built systems that combined classification, entity extraction, and semantic relationships. The interface wasn't the product. The structure of information was. It was defining how discovery itself should work.

Discovery is not about returning results. It's about helping users navigate possibility without getting lost.

Complex Operational Systems

At Tavant, I worked on integrating multiple financial systems into a single browser-based platform—long before "platform thinking" became standard.

The challenge wasn't UI simplification. It was orchestrating complexity across an entire lifecycle — 7 independent systems, multiple user roles, high-stakes decisions. We designed for continuity, not screens.

Complex systems don't need simplification. They need structure, hierarchy, and flow.

Platforms

At Google, I worked on products that expanded the reach of the internet itself — Maps, Search, and emerging market initiatives.

With Map Maker, the question wasn't usability. It was: How do you build a system where users become the infrastructure? Mapping regions with no prior data, enabling contributions across ~190 countries, designing for non-linear addressing systems. The system only worked if people trusted it enough to contribute.

Platforms are not built by features. They are built by participation systems.

Product Strategy

Working with 100+ startups through Accel and Design Ventures, I saw the same pattern repeat — teams solving the wrong problem extremely well.

My role wasn't to design solutions. It was to reframe problems. What are we actually solving? What signals matter? What should we not build? Most product decisions fail not due to execution — but due to misaligned problem definitions.

The hardest product problem is defining the problem.

Ecosystem

At Tata Digital, I worked on building financial services within a larger ecosystem — payments, credit, insurance, and investments.

These are not standalone products. They are interdependent systems held together by trust. Infrastructure defines capability. Regulation defines constraints. Experience defines adoption. Every decision had system-wide consequences.

You don't design fintech products. You design trust systems.

Learning Systems

At a unicorn edtech company, I led product thinking in a high-frequency, high-variance environment where learning outcomes depend on continuous interaction between students, teachers, and content. The product is not a platform. It is a live learning system.

Engagement and outcomes were inconsistent; the real gap was behavior and feedback, not content. We focused on capturing signals (participation, teacher pace, session dynamics), tightening real-time and post-session feedback loops, and adapting session flow—prioritizing interaction-driven learning and teacher augmentation over content delivery alone.

Learning is not delivered. It is continuously adjusted.

Decision Systems

At a leading insurance company, I worked in a category where users have low understanding, decisions are high-stakes, and trust is fragile. The product is not a marketplace. It is a decision system under uncertainty.

Users struggled to understand products, compare options meaningfully, and trust recommendations. We reframed the work as confident decision-making under cognitive and trust constraints: structuring complex attributes, guided flows with progressive disclosure, outcome-oriented comparison over feature lists, and transparency in trade-offs—balancing business goals with clarity and trust.

The goal is not to help users choose faster. It's to help them choose with confidence.