惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

爱范儿
爱范儿
量子位
大猫的无限游戏
大猫的无限游戏
小众软件
小众软件
J
Java Code Geeks
B
Blog
V
V2EX
博客园 - 三生石上(FineUI控件)
Blog — PlanetScale
Blog — PlanetScale
aimingoo的专栏
aimingoo的专栏
Y
Y Combinator Blog
F
Fortinet All Blogs
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
The Cloudflare Blog
A
About on SuperTechFans
D
DataBreaches.Net
阮一峰的网络日志
阮一峰的网络日志
博客园 - Franky
H
Help Net Security
宝玉的分享
宝玉的分享
Martin Fowler
Martin Fowler
酷 壳 – CoolShell
酷 壳 – CoolShell
MongoDB | Blog
MongoDB | Blog
L
LangChain Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Building Evenfield: An AI Homeschool Platform That Never ...
Jonomor · 2026-04-28 · via DEV Community

Jonomor

I built Evenfield because existing educational platforms treat every session like the first session. A student struggles with fractions in September, makes progress through October, then encounters decimals in November — and the AI tutor has no memory of that fraction journey. This fundamental limitation turns personalized learning into a series of disconnected interactions.

Evenfield solves this by maintaining persistent memory across every tutoring session. Built on Anthropic's Claude and powered by our H.U.N.I.E. memory layer, it's an AI-powered homeschool platform that actually remembers each learner's progress, struggles, breakthroughs, and learning patterns.

The Memory Problem in AI Education

Most AI tutoring platforms operate with session-based memory. Each conversation starts fresh. The AI might adapt within a single session, but by the next day, it's forgotten how the student learns best, which concepts they've mastered, and where they consistently struggle.

This creates a broken learning experience. Real teaching requires understanding not just what a student knows today, but how they got there, what methods worked, and what didn't. Memory isn't a nice-to-have feature in education — it's fundamental to how learning actually works.

Architecture and Technical Decisions

Evenfield runs on Next.js with Supabase handling data persistence and Railway managing deployment. The interface uses Tailwind CSS for responsive design across devices. But the core innovation sits in the integration with H.U.N.I.E., our persistent memory system.

Every tutoring session writes to H.U.N.I.E.'s memory layer. When a learner works through algebra problems, struggles with reading comprehension, or shows aptitude for coding concepts, that information persists. The next session begins with full context of previous interactions.

The AI tutor accesses this accumulated knowledge to make informed decisions about pacing, difficulty adjustment, and instructional approach. It knows if a student learns better through visual examples or verbal explanation, whether they need more practice with foundational concepts, or if they're ready to advance.

Fifteen Subjects, Three Learners

The platform covers fifteen subjects including math, coding, financial literacy, entrepreneurship, AI literacy, Spanish, reading, and science. Each subject maintains its own progression tracking while contributing to a unified learner profile.

Currently serving three learners with two instructors, Evenfield differentiates content by age and skill level. The same algebraic concept gets presented differently to a visual learner versus someone who prefers step-by-step logical progression.

For compliance requirements, the system auto-generates quarterly PDF reports documenting progress across all subjects. These aren't generic templates — they reflect actual learning outcomes tracked through persistent memory.

Integration with the Jonomor Ecosystem

Evenfield represents the first H.U.N.I.E. client application, serving as live proof that persistent AI memory transforms tutoring effectiveness. Each learner operates as an agent that writes to H.U.N.I.E. after every session, building a comprehensive educational profile over time.

This integration demonstrates H.U.N.I.E.'s capability beyond simple data storage. The memory layer understands educational context, maintains learning progression timelines, and enables sophisticated queries about student development patterns.

Building for Real Use

I didn't build Evenfield as a demo or proof of concept. This is the education platform my children use daily. Every feature decision gets validated through actual teaching sessions. Every technical choice supports real learning outcomes.

The platform evolved from immediate needs: tracking progress across multiple subjects, maintaining continuity between sessions, and providing the kind of individualized attention that makes homeschooling effective. The persistent memory capability emerged from watching how human tutors build understanding over time — they remember everything about each student's learning journey.

Evenfield proves that AI tutoring can move beyond session-based interactions toward genuine educational relationships. The technology exists to build systems that truly know their learners.

Learn more at https://www.evenfield.io.