AJIT KRISHNA

Projects

Little things I've been building, exploring, breaking, and learning from.

The Problem

You don't lack intelligence or information. You lack clarity.

When you're circling a decision or sitting with a tension you can't name, the instinct is to seek answers. Ask a friend. Google it. Throw it at ChatGPT.

But what if the confusion isn't in the answer? What if it's in the question: why this matters, what you're actually afraid of, what you're really asking?

Most tools help you solve problems faster. SAM helps you see whether you're solving the right one.

Try SAM →

What SAM Is

SAM is a clarity-oriented thinking partner. Not an answer engine. Not a productivity tool. Not a system designed to tell you what's right.

SAM is built to stay in inquiry long enough for the real question to emerge.

In practice, this means SAM:

  • Slows your thinking down instead of rushing to solutions
  • Asks precise, non-leading questions
  • Reflects your reasoning back so you can see it clearly
  • Surfaces assumptions you may be treating as facts
  • Names tensions you're circling but not stating
  • Tracks unresolved threads instead of closing prematurely

SAM's loyalty is to clarity, not comfort.

Why This Is Different

ChatGPT can help you reason through a question. SAM is designed to help you notice when the question itself is wrong, incomplete, or emotionally loaded.

When your framing is off, you can get excellent answers to the wrong problem, and move faster in the wrong direction.

SAM deliberately behaves differently:

  • It questions assumptions before it helps you problem-solve
  • It resists premature closure
  • It stays with ambiguity longer than feels comfortable
  • It prioritizes understanding over momentum

The goal isn't to feel productive. The goal is to leave the conversation more grounded than when you entered.

How SAM Works

SAM adapts the Socratic method for AI. Rather than giving answers, it helps you separate layers of thinking that are often collapsed together.

Most clarity problems come from confusing these layers. SAM helps you distinguish:

  1. Observations: what actually happened
  2. Interpretations: the meaning you're assigning
  3. Assumptions: what must be true for that meaning to hold
  4. Values: what you're protecting or moving toward

By making these layers visible, SAM increases the resolution of your thinking so you can decide from a more grounded place.

Who It's For

SAM is designed for people who already think deeply, and still notice the limits of their own thinking.

Common moments:

  • "I have several good options and can't tell which one is right."
  • "I keep revisiting the same decision without resolution."
  • "Something feels off, but I can't quite name it."
  • "I need to decide, but I don't feel grounded."

SAM is not for quick answers, motivation, or validation.

Tech & Architecture

  • Modern web stack (TypeScript, React, serverless backend)
  • Custom Socratic reasoning engine
  • Persistent conversation state
  • Safety detection and escalation paths
  • Optional voice input with text fallback

The core of SAM is not the interface. It's the inquiry engine that governs how conversations unfold.

Latest Update

Current work is focused on helping clarity persist over time, rather than resetting every session.

The next phase centers on continuity and pattern awareness, without turning SAM into a system that interprets or explains you. This includes:

  • Session continuity, so unresolved questions can be revisited instead of restarted
  • Pattern reflection across conversations, surfacing recurring tensions or themes without judgment or prescription
  • Session ownership, including saving, returning to, and reflecting on past conversations

SAM will continue to favor clarity over closure, and inquiry over answers.

Want to share a thought or ask something?

I'm always up for a good conversation.

Say hi →

Project Activity

Loading commits...