Devplan
Devplan is an AI-native product management platform for software teams that connects information from code, customer feedback, tickets, documents and conversations. It helps teams prioritize product work, create evidence-backed specifications, and track delivery so people and AI agents share context and can act on it.
Facts
- HQ
- Seattle, WA, US
- Sectors
- Artificial Intelligence
- Employees
- 11–50
- Registered name
- Devplan Inc.
- Stage
- Seed
- Founders
- Chris Bee, Anton Safonov
Is this your organization? Claim it
Chris Bee (Devplan) shares a copyable Keep a Changelog template and examples for semantic versioning
Chris Bee explains the most common changelog format—Keep a Changelog paired with semantic versioning—and provides a copyable Markdown template, examples, and guidance for writing entries from merged work rather than memory. Devplan’s changelog view captures merged code and pull requests to help generate those entries automatically.
Chris Bee: For Reliable Estimates, Size Software Work From the Code — Devplan Maps Files to Day Ranges
Chris Bee argues that accurate software estimates require examining the code and using ranges plus comparisons to past projects rather than single-number guesses made in meetings; Devplan automates sizing by mapping changes to files and services and producing lead-reviewed day ranges.
Chris Bee explains spec-driven development and how Devplan supplies specs to Claude Code via MCP
Chris Bee explains spec-driven development (SDD) as writing and agreeing a specification before any AI coding agent generates code, so the spec—containing goals, requirements, edge cases, and acceptance criteria—serves as the source of truth; Devplan generates such specs and feeds them to Claude Code via MCP. The post outlines SDD phases (Specify, Clarify, Plan, Tasks, Implement, Check) and contrasts SDD with ad-hoc "vibe coding," using a CSV export example and noting compatibility with plain Markdown.
Chris Bee explains AI product management and how Devplan automates request gathering, spec writing, estimates, and tracking
Chris Bee explains AI product management as using AI to handle evidence gathering and administrative product tasks—collecting and ranking customer requests, writing specs from code, estimating work, and tracking status—while the human product manager focuses on decisions; Devplan is presented as an AI-native tool that does this across existing tools.
Chris Bee (Devplan) lists best AI tools for product managers in 2026 and positions Devplan as the integrator
Chris Bee outlines the best AI tools product managers should use in 2026 for specific jobs—research, specs, planning, meetings, tracking and prototyping—and explains that Devplan connects these siloed tools so customer requests, specs, code and status live in one place.
Devplan’s Live Docs automatically writes and updates product documentation from your codebase
Devplan's Live Docs is an AI feature that generates and updates product documentation directly from a codebase, creating a page for every detected feature and keeping docs in sync as code changes. It aims to reduce time spent tracking down behavior by giving support, sales, PMs, and engineers a single source of truth derived from GitHub, GitLab, or Bitbucket.
Devplan publishes blog focused on product management, context, and AI coding
Devplan's blog covers product management, contextual considerations, and AI coding, offering first-party perspectives from the Devplan team.
Devplan Team explains how to connect Claude Code to Devplan MCP so Claude checks customer requests and decisions before coding
Devplan’s guide shows how to connect Claude Code to product context using one command and a short CLAUDE.md update so Claude can check customer requests, past decisions, and project status before writing code.
Chris Bee’s Six-Step Guide: Turn Customer Feedback into a Buildable PRD with Template and Example
Chris Bee outlines a six-step method to convert customer feedback into a buildable PRD, starting from collected evidence, defining the problem in customers' words, setting a single measurable goal, and listing prioritized requirements with clear acceptance criteria and handoff-ready details.
Chris Bee evaluates 9 Jira alternatives and recommends keeping Jira for tickets while adding a tool for product decisions
Chris Bee reviews nine Jira alternatives for software teams, explaining which tools suit fast engineering teams, cross-functional work, self-hosting, and Microsoft shops, and proposes keeping Jira for tickets while adding a separate tool to decide what to build.
Chris Bee compares Linear and Jira: Linear for fast software teams, Jira for configurable enterprise planning
Chris Bee compares Linear and Jira across setup, workflows, planning, integrations, AI, and pricing, recommending Linear for fast, opinionated software teams and Jira for organizations needing deep configuration and cross-team planning. He notes both tools track work well but neither helps decide what to build from customer evidence — a gap Devplan aims to fill.
Chris Bee’s guide: Product managers can use Claude Code without engineering skills to analyze code, customer feedback, and prototype
This guide explains how product managers can use Claude Code without coding skills to investigate codebases, analyze customer feedback, draft PRDs, and prototype, emphasizing that value depends on supplying product and customer context (e.g., via Devplan).
Chris Bee compares Dovetail to Condens, Looppanel, Aurelius, Marvin, Great Question, Devplan, Productboard, Enterpret and Canny — recommends research repositories for research programs and...
Chris Bee compares Dovetail with alternatives for product teams, arguing that research repositories (Condens, Looppanel, Aurelius, Marvin, Great Question) suit research programs while feedback-to-roadmap tools (Devplan, Productboard, Enterpret, Canny) better serve teams needing customer feedback turned into a roadmap, and highlights pricing and workflow fit as key decision factors.
Chris Bee explains a six-step method to convert customer feedback from calls, tickets and Slack into a prioritized product roadmap
Chris Bee outlines a six-step process to consolidate customer feedback from calls, tickets, and Slack into a single list, deduplicate and score requests by number of customers and revenue impact, rank them against strategy and effort, write specs for prioritized items, and communicate outcomes to customers; Devplan automates these steps but the method works with spreadsheets too.
Chris Bee shares six annotated PRD examples from Devplan’s free template covering web, mobile, backend, payments, healthcare, and marketplace
Chris Bee provides six annotated PRD examples (web app, mobile app, backend service, payments, healthcare, marketplace) drawn from Devplan’s free PRD template, showing what a strong PRD includes: customer problem and evidence, a measurable goal, prioritized requirements, scope exclusions, and testable acceptance criteria.
Devplan Team: Company Context Should Compound — Shared MCP Connections Turn Individual Fixes into Reusable Organizational Knowledge
Devplan argues that giving employees access to company data (via MCP) speeds individual work but without a shared context layer the same conflicts and decisions are repeatedly resolved by individuals rather than captured as organizational knowledge. A governed, reusable context layer would turn solitary corrections into company-wide learning and reduce recurring interruptions and inconsistency.
Devplan Team outlines automated Claude workflow to produce verified weekly product status reports across GitHub, Linear/Jira, and Slack
Devplan’s guide explains how to automate a trustworthy weekly product status report in Claude that reviews GitHub, Linear/Jira, and Slack, verifies claims against sources, flags blockers/decisions/risks, and runs every Monday morning.
Devplan Team says Claude Code and similar coding agents can't reliably build or maintain a company's knowledge graph
Devplan argues that coding agents like Claude Code can search and act across company systems but are insufficient for building and maintaining an organizational knowledge graph because they can't keep its continuously updated, evidence-linked model of relationships and context correct as organizations change.
Devplan Team: Companies Need a Knowledge Graph to Turn AI Tool Connections into Organizational Intelligence
Devplan argues that direct AI connections to tools like GitHub, Jira, and Slack aren’t enough for organizational intelligence; companies need a knowledge graph to create shared memory, governance, and reusable context so leaders can answer higher-level questions about how the organization actually works.
Anton Safonov’s Devplan unveils Weaver context graph to provide trusted, evidence-backed product knowledge
Anton Safonov explains Devplan’s context graph, Weaver, as a structured product knowledge graph designed to connect people, projects, documents and decisions so humans and AI can discover trustworthy, evidence-backed, and context-aware information tailored to different roles.
Devplan Team: AI Speeds Development, Shifts Product Management Toward Coordination of Strategy, Execution, and Knowledge
Devplan Team argues that AI has accelerated software development so much that the main challenge now is organizational coordination: deciding what to build, keeping shared understanding, and aligning human and AI agents rather than writing code faster.
Devplan Team: High-performing AI Teams Win by Engineering Context, Not Better Prompts
The article argues that high-performing AI teams succeed not by writing better prompts but by engineering richer context—structuring memory, feedback, and relevant artifacts so models receive the precise information they need to act reliably.
Devplan Team: Environment Design, Not Model Choice, Determines AI Coding Agent Success
Devplan argues that AI coding agent failures stem from poor environment design—not model choice—and explains that a robust harness (tools, context management, memory, guardrails, and verification) is essential to make agents reliable.
Chris Bee: AI Speeds Coding but Lost Intent Becomes Development’s New Bottleneck
Chris Bee argues that AI accelerates coding but loses developer intent in handoffs, producing technically working but misaligned, hard-to-maintain solutions; the real bottleneck is clarifying intent and inputs, not execution speed.
Chris Bee: Devplan Powers AI-Assisted Development by Scoping Specs, Breaking Work into Half-Day to 5-Day Features, and Automating Execution
Chris Bee explains how Devplan structures AI-assisted software development into scoped specs, right-sized features, and automated workflows so AI can independently deliver work in parallel with minimal human oversight. The post argues Devplan reduces management overhead and makes over 90% of shipped code run through a repeatable, scalable system.
People
-
Anton Safonov Co-founder and CTO
-
Chris Bee Co-founder and CEO