Sign in Subscribe

Drishti Shah

How to host an AI Hackathon without losing control of your keys or budget

How to host an AI Hackathon without losing control of your keys or budget

Running an AI hackathon means providing dozens (or hundreds) of teams with access to expensive LLM APIs while maintaining cost control, fair usage, and visibility into what's happening. Without proper infrastructure, you risk: - Budget blowouts: A single runaway script consuming your entire API budget- - - No

The best approach to compare LLM outputs

The best approach to compare LLM outputs

Once LLMs are in production, output quality stops being a subjective question and becomes an operational one. Teams are no longer asking whether they need to evaluate outputs, but how to do it reliably as systems evolve. Production systems can change frequently. Prompts are iterated on, models are swapped, routing

OpenAI Responses API vs. Chat Completions vs. Anthropic Messages API

OpenAI Responses API vs. Chat Completions vs. Anthropic Messages API

A side-by-side comparison of OpenAI's Chat Completions, Responses API, and Anthropic's Messages API, covering key differences, use cases, and how to avoid vendor lock-in with Portkey.

Introducing the MCP Gateway

Introducing the MCP Gateway

Your interns need three approvals to touch production. Your AI agents? Zero. With MCP, agents can take real action – connect to databases, trigger workflows, access internal systems. The protocol just works. But here's what we kept hearing from teams actually running MCP in production: "How do I

LLM hallucinations in production

LLM hallucinations in production

Hallucinations in LLM applications increase at scale. This blog explains how AI gateways and guardrails help control, detect, and contain hallucinations in production systems.

OpenCode: token usage, costs, and access control

OpenCode: token usage, costs, and access control

A practical guide to managing OpenCode token usage, costs, and governance in shared or production environments.

MCP tool discovery for autonomous LLM agents

A concise summary of the MCP-Zero paper, explaining how active tool discovery enables scalable, autonomous LLM agents while reducing context overhead.