The Apptension blog
Written by the engineers, designers and product leads who do the work. How we scope MVPs, pick a stack, run AI-assisted delivery and look after products once they are live.
Newest first
323 articles · page 2 of 27
Process14 min read
Many agentic MVPs die before the first sprint because the ask is too broad, too mobile heavy, or too risky to ship safely. This article walks through Apptension’s fit gate with concrete scope examples for B2B SaaS, workflow products, and internal tools.
Apptension Team
Business13 min read
July 2026 signals from Hacker News, Product Hunt, and GitHub trending point to a clear pattern: teams want AI coding agents, but they do not want to bet the org on one vendor. The practical response is a “harness” approach: standard prompts, policies, evals, and telemetry that work across tools like OpenCode, Aider, Claude Code, and Copilot CLI.
Apptension Team
Business16 min read
A first MRR dip is rarely a glitch. It is usually flat acquisition plus rising churn catching up. This article shows how to diagnose the break (channel to expansion) and run focused experiments instead of panic shipping.
Apptension Team
Process12 min read
A 2026 playbook for SaaS product managers building in an AI-heavy landscape. Focuses on discovery, measurable outcomes, delivery flow, and risk controls that hold up in production.
Apptension Team
QA13 min read
QA changes when your product includes an LLM or ML model. This article shows how to test AI behavior with datasets, judges, automation, and drift monitoring - without pretending outputs are deterministic.
Apptension Team
QA12 min read
AI changes what “quality” means. You still test the app, but you also test data, prompts, models, and outcomes. This article shows practical QA patterns, metrics, and CI checks that work for AI features.
Apptension Team
QA14 min read
AI features break familiar QA assumptions: outputs vary, models drift, and bugs look like “bad answers.” This article shows how we test AI systems in delivery work, with concrete checks, tooling, and automation patterns.
Apptension Team
Technology14 min read
AI help is now part of daily engineering work, but it still breaks in predictable ways. This article covers practical workflows, guardrails, and tool patterns that hold up in production in 2026.
Apptension Team
Technology11 min read
AI coding tools work well for some developers and fall apart in shared codebases without guardrails. This article explains how to move from experiments to team-wide adoption with measurable quality and delivery outcomes.
Apptension Team
Technology11 min read
AI coding tools can speed up routine work, but they also create new failure modes. This article shows what actually works in day to day delivery, and how to set guardrails that keep quality and security intact.
Apptension Team
Technology13 min read
AI coding assistants can speed up delivery, but they also amplify security and maintenance risks. This article maps what it takes to move from early adopters to the early majority: proof, controls, and repeatable guardrails.
Apptension Team
Technology15 min read
SaaS security is a chain of small controls that fail in predictable ways. This guide covers the essentials: identity, secrets, data protection, app hardening, monitoring, and incident response with practical implementation details.
Apptension Team