Smart Tools, Human Touch: Using AI in Product Without Losing the Plot
Smart Tools, Human Touch: Using AI in Product Without Losing the Plot
I work with AI-powered knowledge management tools every day, Glean, JPD, and the wider stack that helps a team find what it already knows. I am genuinely enthusiastic about them. I am also clear that they are there to make people more effective, not to replace the judgement that makes a product good.
The best tool problem is a findability problem
Most teams do not suffer from a lack of information. They suffer from not being able to find the decision they made six months ago, or the requirement that already answers today's question. That is where AI-powered knowledge management actually shines: surfacing what already exists so we stop solving the same problem twice.
When discovery is faster because context is one search away, the whole team spends more time on the questions that genuinely need a human.
Use AI for the draft, not the decision
I lean on smart tools to accelerate the parts that benefit from speed, first drafts, summaries, pulling threads together from scattered notes. What I do not outsource is the decision. Prioritisation, trade-offs, and reading what a stakeholder actually means still belong to a person who understands the context.
The tools give me a faster start. They do not give me the judgement about whether something is the right thing to build.
Keeping the human touch central
The reason I care about this balance is simple: every sprint, every backlog item, every requirement traces back to a real person trying to get something done. AI can help me move faster, but it cannot feel the user's frustration or sense when an alignment conversation is quietly going sideways.
So I treat smart tools as exactly that, tools. They make teams more effective and let me connect strategy to execution with less friction. But the human touch stays central to every decision, because that is still where good products come from.
Tanvi skipped presentations and built real AI products.
Tanvi Juneja was part of the April 2026 cohort at Curious PM, alongside 18 other talented participants.