jhol.dev
Jhol Hewres is an independent principal software engineer and open-source creator focused on AI agents, developer tooling, and production-ready AI systems. His stated philosophy is to ship reliable AI with observability, cost control, latency awareness, and graceful degradation rather than prototypes alone.
At a Glance
- Individual software developers and AI-assisted coding practitioners
- Engineering teams building or operating AI agents
- Teams that use multiple AI coding tools and need shared project memory
- Hospitality, real-estate, and legal-service businesses using WhatsApp automation
- +1 more
AI Tools by jhol.dev
(1)
devpit
Open Source Coding Agent Workspace
Discussions
No discussions yet
Be the first to start a discussion about jhol.dev
Latest News
Products & Services
A native, free and open-source desktop app for controlling coding agents: one workspace per project with terminal, board, chat/orchestrator, sessions, costs, files/diffs, browser, notes, and Excalidraw. It keeps workspace data outside the repository in ~/.devpit and currently drives Claude Code end to end.
A local-first MCP server in Go that provides persistent, cross-tool memory for Claude Code, Cursor, OpenCode, Gemini CLI, Codex, VS Code Copilot, and other MCP-compatible tools. It combines hybrid vector/BM25 search, local ONNX embeddings, sanitization, SQLite, and a bitemporal knowledge graph.
A self-hosted team-memory server that syncs project facts, decisions, and learnings across machines and tools, with privacy guardrails, audit logging, an embedded dashboard, and Postgres or SQLite support.
An open-source AI agent for tech teams distributed as a single Go binary, with CLI, WebUI, MCP server, WhatsApp and Telegram interfaces, persistent memory, encrypted credential vault, multiple LLM providers, and 90+ built-in tools.
Market Position
Hewres positions his projects as lean, local-first alternatives to heavier or siloed AI tooling: DevClaw emphasizes a secure single Go binary versus larger agent stacks such as OpenClaw; Anchored addresses the tool lock-in and fragmented memory found across Claude Code, Cursor, OpenCode and similar tools; and devpit frames itself as an Agentic Development Environment centered on the project and workflow rather than as another conventional IDE. Relevant comparison points include OpenClaw, Claude Code, Cursor, OpenCode, and conventional IDE-based agent products.
Leadership
Founders
Jhol Hewres
Principal Software Engineer specializing in AI agents, developer tooling, and AI infrastructure. His background includes 10+ years in software engineering, leading teams of up to 11 engineers, building financial systems and platforms serving 100k+ users, co-founding Alta Forja Tecnologia, and co-founding IntegraBot.ai.
Founding Story
jhol.dev is a personal portfolio and publishing identity rather than a separately identified corporation. Hewres uses it to publish his work and writing and to develop open-source AI and developer tools. Individual projects arose from practical problems in his own engineering workflow: DevClaw from wanting a leaner, more secure coding agent; Anchored from memory being trapped inside individual AI coding tools; and devpit from plans, reviews, research, and project artifacts being scattered across invisible directories and applications.
Business Model
Revenue Model
The public portfolio projects are presented as open-source or free/local-first software. devpit is explicitly free under Apache 2.0; its account is optional and currently only supports sign-in. IntegraBot.ai and IntegraClaw are described as SaaS products with agent integrations, billing/rate limiting, and management features.
Pricing Tiers
Open source under Apache 2.0; runs on the user's machine and uses the user's existing Claude Code plan rather than requiring new API keys.
Target Markets
- Individual software developers and AI-assisted coding practitioners
- Engineering teams building or operating AI agents
- Teams that use multiple AI coding tools and need shared project memory
- Hospitality, real-estate, and legal-service businesses using WhatsApp automation
- Developers and organizations preferring local-first or self-hosted AI tooling
- AI-assisted software development, code review, testing, and deployment
- Managing multiple coding-agent sessions and projects in a local Agentic Development Environment
- Sharing durable project memory across AI coding tools and development teams
- Building autonomous and collaborative multi-agent applications
- WhatsApp-based customer support and business-process automation
- Self-hosted AI infrastructure where credentials, data, and memory remain under the developer or team's control