DEV Community

Amit profile picture

Amit

I run agents as daily infrastructure — not demos. Artificial Curiosity Labs is where I document what actually works: agent fleets, skills as institutional memory, cost discipline, and the gap between

Most Enterprises Are Chasing AI Autonomy Before Defining the Outcome

Most Enterprises Are Chasing AI Autonomy Before Defining the Outcome

Comments
9 min read
Wiring Codex and ChatGPT Desktop to OpenAI Models on Amazon Bedrock Runtime

Wiring Codex and ChatGPT Desktop to OpenAI Models on Amazon Bedrock Runtime

Comments
5 min read
Alternative AI Hardware Is a Systems Problem

Alternative AI Hardware Is a Systems Problem

Comments
7 min read
SambaNova Compiles Models Into a Rack

SambaNova Compiles Models Into a Rack

Comments
9 min read
The GPU Stopped Being the Product

The GPU Stopped Being the Product

Comments
16 min read
There Is No Universal AI Runtime—So Build a Portable Control Plane

There Is No Universal AI Runtime—So Build a Portable Control Plane

Comments
9 min read
The Interconnect Determines How Much of the Chip You Can Use

The Interconnect Determines How Much of the Chip You Can Use

Comments
11 min read
Groq Makes the Compiler Part of the Processor

Groq Makes the Compiler Part of the Processor

Comments
9 min read
d-Matrix Moves the Math Into Memory

d-Matrix Moves the Math Into Memory

Comments
7 min read
When the Platform Operator Designs the Chip

When the Platform Operator Designs the Chip

Comments
9 min read
Cerebras Moved the Cluster Boundary Onto a Wafer

Cerebras Moved the Cluster Boundary Onto a Wafer

Comments
11 min read
AMD Is Rebuilding the GPU Stack in the Open

AMD Is Rebuilding the GPU Stack in the Open

Comments
7 min read
The AI Supply Chain Became Part of the Architecture

The AI Supply Chain Became Part of the Architecture

Comments
10 min read
The AI Rack Became a Data Center Design Problem

The AI Rack Became a Data Center Design Problem

Comments
13 min read
The AI Buildout Is Constrained by Its Slowest Layer

The AI Buildout Is Constrained by Its Slowest Layer

Comments
12 min read
One Coding Agent, Three Bedrock APIs: Wiring OpenCode to Claude, GPT, Grok, and Kimi

One Coding Agent, Three Bedrock APIs: Wiring OpenCode to Claude, GPT, Grok, and Kimi

Comments
8 min read
Jev Does Not Write. That Is the Point.

Jev Does Not Write. That Is the Point.

Comments
6 min read
So What Is the Effective Cost for a Million Tokens on Frontier Models?

So What Is the Effective Cost for a Million Tokens on Frontier Models?

Comments
5 min read
Search Is a Pipeline, Not a Tool

Search Is a Pipeline, Not a Tool

Comments
6 min read
A 200 Response Does Not Prove Search Worked

A 200 Response Does Not Prove Search Worked

Comments
8 min read
Search for Agents Is Becoming Its Own Market

Search for Agents Is Becoming Its Own Market

Comments
7 min read
What Actually Differs Across Search APIs Built for Agents

What Actually Differs Across Search APIs Built for Agents

Comments
5 min read
Open Weights Catch the Last Frontier, Not the Moving One

Open Weights Catch the Last Frontier, Not the Moving One

Comments
8 min read
The Gateway Controls the Request. The Router Chooses the Model.

The Gateway Controls the Request. The Router Chooses the Model.

Comments
9 min read
The Agentic DevSecOps Stack Is Taking Shape

The Agentic DevSecOps Stack Is Taking Shape

Comments
11 min read
Exploring the Voice AI Stack

Exploring the Voice AI Stack

Comments
10 min read
An Agent That Can Pay for Its Tools

An Agent That Can Pay for Its Tools

Comments
9 min read
MPP vs x402: Two Competing Answers to the Same HTTP 402

MPP vs x402: Two Competing Answers to the Same HTTP 402

Comments
4 min read
A Browser That Pays

A Browser That Pays

Comments
4 min read
AgentCore Registry Is a Governed Catalog for Agent Interfaces

AgentCore Registry Is a Governed Catalog for Agent Interfaces

Comments
6 min read
What the Trace Knew That the Logs Didn't

What the Trace Knew That the Logs Didn't

Comments
11 min read
Multi-Agent Memory: Agents That Actually Share Context

Multi-Agent Memory: Agents That Actually Share Context

Comments
6 min read
Beyond Semantic Retrieval: Metadata Filtering in Agent Memory

Beyond Semantic Retrieval: Metadata Filtering in Agent Memory

Comments
5 min read
Bedrock Managed Knowledge Bases as a Native Gateway Tool

Bedrock Managed Knowledge Bases as a Native Gateway Tool

Comments
5 min read
Harness vs. Runtime: When to Graduate

Harness vs. Runtime: When to Graduate

Comments
7 min read
Build a Regression Suite for an Agent

Build a Regression Suite for an Agent

Comments
6 min read
AgentCore Browser: A Managed, Isolated Browser an Agent Can Drive

AgentCore Browser: A Managed, Isolated Browser an Agent Can Drive

Comments
8 min read
Grok 4.6 Is on Amazon Bedrock. Here Is How to Wire Grok Build to It

Grok 4.6 Is on Amazon Bedrock. Here Is How to Wire Grok Build to It

Comments
7 min read
Where AgentCore Sits in the Architecture

Where AgentCore Sits in the Architecture

Comments
7 min read
What You Build on the AgentCore Harness

What You Build on the AgentCore Harness

Comments
6 min read
The AgentCore Map: What Each Piece Actually Does

The AgentCore Map: What Each Piece Actually Does

Comments
10 min read
Who May Call What: Per-User Authorization on AgentCore With Cedar

Who May Call What: Per-User Authorization on AgentCore With Cedar

Comments
8 min read
Browser Use vs. AgentCore Browser: Two Managed Browsers, and Which Layer Each One Wins

Browser Use vs. AgentCore Browser: Two Managed Browsers, and Which Layer Each One Wins

Comments
9 min read
Two Ways to Authorize an Agent Tool: IAM or OAuth on Amazon Bedrock AgentCore Gateway

Two Ways to Authorize an Agent Tool: IAM or OAuth on Amazon Bedrock AgentCore Gateway

Comments 2
7 min read
AgentCore Runtime: Where an Agent Actually Runs

AgentCore Runtime: Where an Agent Actually Runs

Comments
5 min read
AgentCore Memory: What an Agent Remembers When the Session Is Gone

AgentCore Memory: What an Agent Remembers When the Session Is Gone

Comments
10 min read
Four Tools, One Loop: The AgentCore Knowledge Worker

Four Tools, One Loop: The AgentCore Knowledge Worker

Comments
8 min read
AgentCore Code Interpreter: When the Agent Has to Actually Run the Math

AgentCore Code Interpreter: When the Agent Has to Actually Run the Math

Comments
7 min read
Benchmark saturation is real. The price collapse is an illusion.

Benchmark saturation is real. The price collapse is an illusion.

Comments
3 min read
Dictation and Meeting Notes, Fully Local: Rebuilding Wispr Flow and Granola Without the Cloud

Dictation and Meeting Notes, Fully Local: Rebuilding Wispr Flow and Granola Without the Cloud

1
Comments 1
6 min read
Building Voice Agents with LiveKit, Cartesia, and AWS

Building Voice Agents with LiveKit, Cartesia, and AWS

Comments
6 min read
The AI Chips After NVIDIA

The AI Chips After NVIDIA

Comments
12 min read
The Five-Layer AI Stack Is an Inverted Pyramid

The Five-Layer AI Stack Is an Inverted Pyramid

Comments
8 min read
MCP Became Stateless: Migrating a Real AgentCore Gateway

MCP Became Stateless: Migrating a Real AgentCore Gateway

Comments
6 min read
Building an Autonomous AI News Agent on Step Functions and AgentCore

Building an Autonomous AI News Agent on Step Functions and AgentCore

Comments
9 min read
Open Data Wasn't Missing. The Interface Was.

Open Data Wasn't Missing. The Interface Was.

Comments
6 min read
Every Way to Attribute Cost on Amazon Bedrock, Live-Tested

Every Way to Attribute Cost on Amazon Bedrock, Live-Tested

Comments
8 min read
Web Search as a Managed Connector: Wiring Amazon Bedrock AgentCore WebSearch Into Any Agent

Web Search as a Managed Connector: Wiring Amazon Bedrock AgentCore WebSearch Into Any Agent

Comments
9 min read
Does Graphify Actually Help an AI Coding Agent?

Does Graphify Actually Help an AI Coding Agent?

Comments
9 min read
Prompt Caching Is a Harness Design Problem, Not an API Toggle

Prompt Caching Is a Harness Design Problem, Not an API Toggle

Comments
7 min read
loading...