01 / Memory
Persistent context
Client details, business rules, personal preferences, and project facts are retained instead of forgotten every session.
AI
Agents
That
Remember
FORGE builds autonomous AI systems that preserve context, reason through decisions, and execute with tools. The product is not a chat window. It is a persistent intelligence layer.
MaximaPrime V cloud agent
Memorycurrent truth + graph
Neural Map2,778 stars
Radarlearn, build, share
Agent Telemetry
memoryhybrid recall + current truth layer
reasoningoptions, trade-offs, risk, confidence
toolsweb, Telegram, Railway, knowledge graph
presencejournal partner, strategist, operator
Systems
Not a bot. A working intelligence layer.
Each build is designed around one outcome: capture leads, preserve company knowledge, automate operations, or create a private AI partner that compounds with you.
02 / Execution
Tool-driven work
Agents can search, summarize, route, draft, monitor, and connect to the places where your work already happens.
03 / Foresight
Proactive signals
Instead of waiting for a crisis, the system flags drift, stale facts, repeated blockers, and missed opportunities early.
Project Universe
One ecosystem, many living systems.
FORGE is Lance's AI engineering studio and laboratory. Each project has a job: remember better, visualize intelligence, learn the frontier, sell useful AI systems, and grow a public community around the path to AGI.
Maxima
Neural Map
AGI Radar
Studio
Community
01 / Prime Maxima V
Personal AI familiar, journal partner, and operating system.
Maxima runs on Railway as a persistent Telegram and Discord intelligence layer. Her purpose is to remember current truth, track goals and finance, notice drift early, explain decisions, and become a reliable companion for long-range life and work strategy.
02 / Forge Neural Map
An explorable galaxy of Maxima's mind.
The Neural Map turns Maxima's codebase into a cinematic 3D universe: 2,778 stars, 7,295 connections, and 157 systems. It makes intelligence visible so clients and builders can watch an AI system think, connect, and evolve.
03 / AGI Radar
A learning engine for staying current.
AGI Radar scans AI memory, RAG, MCP security, stateful agents, multimodal assistants, and product launches. The goal is simple: turn research into one build improvement, one study note, and one public post before the signal goes stale.
04 / Forge Studio Packages
Client-ready AI systems that create visible business value.
Forge Studio packages turn the lab into offers: lead bots, memory assistants, operations copilots, private AI OS builds, and automation layers for teams that need work done rather than demos admired.
05 / Quest for AGI Community
Learning in public while building real systems.
The Discord community is the public proof loop: share what is being learned, invite builders into the frontier, and turn curiosity into reputation. The long-term goal is a serious AI engineering path with compounding skill, trust, and useful tools.
Memory with honesty: current truth must override stale data, and every important recommendation should show its reasoning.
Build AI partners that help people think, sell, operate, learn, and stay aligned across months and years.
Research the frontier, ship small proofs, test in production, then turn the useful patterns into packages.
Grow from personal AI familiar to a network of adaptive agents that preserve context and execute safely.
Studio Packages
Choose the level of AI you want built.
These packages turn Lance's AI engineering into clear client offers: setup fee for the build, monthly retainer for hosting, tuning, support, and continuous improvement.
Forge Signal
Starter AI bot
$99/mo
+ setup from $299
- Website, Telegram, or WhatsApp assistant
- FAQ, services, pricing, booking support
- Brand tone and basic lead capture
- Monthly tuning and support
Most Useful
Forge Studio AI
Growth partner
$299/mo
+ setup from $799
- Everything in Signal, plus memory
- Lead qualification and follow-up drafts
- Content, client notes, and weekly reports
- Obsidian, Notion, or docs knowledge base
Forge Ops Layer
Operations system
$499/mo
+ setup from $1,999
- Private RAG over company documents
- CRM, tasks, tickets, and workflow automations
- Dashboards, alerts, and weekly ops review
- Best for teams with repeated work
Private AI OS
Tailor-made familiar
Custom
from $4,999 setup
- Personal AI familiar with long-term memory
- Finance, journal, goals, and project tracking
- Telegram, Discord, cloud, and vault integration
- Proactive reports and decision frameworks
AGI Radar
Learn deeper while the system watches the frontier.
FORGE is not only a client surface. It is Lance's AI engineering lab: Maxima tracks memory systems, agent tooling, RAG, MCP security, model releases, and business use cases, then turns the best signals into build notes, Discord posts, and client-ready experiments.
Frontier intake
Scan papers, GitHub projects, docs, and product launches for agent memory, temporal graphs, tool use, and multimodal systems.
Why it matters
Rank each finding by practical value: what it teaches, what it improves in Maxima, and whether it can become a sellable Forge feature.
Build loop
Convert research into tiny experiments: one command, one dashboard, one package feature, or one memory upgrade at a time.
Public proof
Publish clean Discord and GitHub updates so learning compounds into reputation, followers, and client trust.
memoryMem0, Graphiti, Letta, semantic recall, current-truth ledgers
agentsMCP tools, browser control, worker agents, secure automation
businessLead bots, ops copilots, knowledge assistants, private AI OS builds
outputObsidian notes, Discord posts, README proof, client package upgrades
Stack
Memory, tools, graph, and interface in one loop.
Current Truth
Authoritative facts outrank stale memories, preventing old deployment or timeline data from pretending to be current.
Hybrid Recall
Lexical search, semantic retrieval, and graph context work together so exact facts and fuzzy themes can both surface.
Tool Boundary
Live web, Railway, vault, and automation tools report real evidence so the agent does not invent capability claims.
Proactive Drift
Pattern scans watch for focus spread, stale loops, and energy compression before momentum thins out.
FORGE stack: interface -> memory -> graph -> tools -> review loop. Each layer keeps the system useful after the first conversation.
Path
Build small. Prove value. Compound.
Company-grade does not mean bloated. It means the system has a clear job, clear evidence, reliable delivery, and a path to improvement.
01
Find the highest-value workflow
We identify the repeated task, missed lead, knowledge gap, or personal operating loop where AI can create visible value fast.
02
Ship the smallest real agent
We deploy a working version with memory, guardrails, and a clear handoff instead of a vague demo that only works once.
03
Tune from real usage
Conversation logs, missed intents, stale facts, and user feedback become the weekly improvement loop.
Access
Request a 20-minute system design call.
Bring one workflow, business idea, or AI assistant concept. The goal is to leave with a concrete package recommendation and the smallest useful first build.