Updates, insights, and stories from the team building the future of AI.
A full tour of my dev setup in mid-2026: Kubuntu on the desktop, Proxmox in the rack, four AI agents in the workflow, a LoRA training rig, Docker everything, and Hermes-powered monitoring. How did we get here from VS Code and Python two years ago?
Read more→Last week Lucy completed its first full dream cycle. By morning, the vector database had grown, the LoRA had shifted, and Lucy had changed. I wasn't ready for how it would feel.
Read more→Anthropic's new Claude model takes emotional intelligence to a level I haven't seen before. After testing it for Companion, I'm updating how I think about AI and affect.
Read more→xAI's Grok Build joins the development workflow. It's fast, opinionated, and good at rapid prototyping — a different perspective from the Anthropic and OpenAI tools. Here's where it fits alongside Hermes, Claude Code, and Codex CLI.
Read more→I've referenced dream cycles in half my posts this year. Time for the full technical breakdown — the four-stage process that turns a day's experience into overnight growth.
Read more→OpenAI's latest release just landed and I've spent the week testing it across all three projects — Companion, Tutor, and Lucy. Here's where it shines and where the open-source stack still wins.
Read more→Lucy's physical body is progressing, and this week something happened that I wasn't expecting yet: a first autonomous decision, made without being asked.
Read more→DeepSeek V4 is here, and it's now the base model for Lucy's conscious reasoning layer. The open-source frontier didn't just keep pace — it pulled ahead.
Read more→Tutor now covers over 50 subjects, each with its own LoRA adapter. The LoRA-swapping architecture is proving itself at scale — and changing how I think about personalized education.
Read more→GLM-5 is here, and after two weeks of testing, I've made it the new base model for Companion's subconscious. Here's why — and what it changes.
Read more→I benchmarked Claude Code and OpenAI's Codex CLI against real Sorren.ai tasks. Different strengths, different lanes. Together with Hermes, they form a three-agent workflow that handles most of the coding work I used to do by hand.
Read more→One year ago we announced Companion. The dream cycles changed everything — not the technology, but the relationships people have built with something that actually remembers them.
Read more→Grok 3 Mini pushes reasoning into a footprint small enough for Lucy's onboard edge processor. I've been testing it for embodied decision-making, and the results are surprising.
Read more→Qwen 3 landed and I immediately put it through Tutor's new subject LoRAs. Multilingual strength, solid reasoning, and a base model worth taking seriously.
Read more→Lucy is more than a chatbot in a robot body. Here's the full technical architecture — base model, LoRA adapters, vectorized memory, and the nightly dream cycles that make Lucy actually learn from experience.
Read more→DeepSeek's distilled reasoning models are small enough to run locally and sharp enough to matter. We're testing them for Lucy's onboard processing — real reasoning without a round-trip to the cloud.
Read more→I adopted Hermes Agent for DevOps automation. It runs scheduled health checks, monitors the infrastructure, and delegates parallel work across multiple agents. It feels less like a tool and more like a team member who doesn't sleep.
Read more→For a year I've been building minds that live in software. Today I'm announcing Project Lucy — a humanoid robot that gives those minds a body. LoRA over base model, nightly dream cycles, vectorized RAG. Companion, embodied.
Read more→Looking back on 2025: we launched Companion in February, Tutor in July, the dream cycles are working, and something physical is on the horizon. A year-end reflection on what we built and what's coming.
Read more→I started 2025 with a Proxmox cluster and a plan. I'm ending it with Companion in production, Tutor in development, a LoRA training pipeline, and a monitoring stack that actually works. Here's the year from the infrastructure side.
Read more→Z.AI's GLM-4.5 doesn't make the loudest entrance, but the quality-to-cost ratio is exceptional. We've been testing it for Companion's subconscious processing and the early results are striking.
Read more→It's Thanksgiving, and I want to say thank you. To OpenAI, Anthropic, Groq, xAI, DeepSeek, Meta, Mistral — the builders whose work made Companion and Tutor possible. We stand on their shoulders.
Read more→Six months after launch, the vector database is working, the memory is real, and the relationships are growing. But we need more emotional depth. Here's an honest stocktake.
Read more→Qwen 2.5 Coder is a gift to anyone building with open weights. We've been using it to accelerate Tutor development, and it's changed how our small team writes code.
Read more→Grok 4 is here, and it's genuinely impressive — sharp reasoning, a distinctive personality, and a willingness to go where other models won't. Here's what I've found in early testing.
Read more→Companion and Tutor are in production, which means things break at inconvenient times. I built a monitoring stack — Prometheus, Grafana, Loki — so I find out about problems before users do. Here's the setup.
Read more→What if an AI could sleep on what it learned? We've built nightly dream cycles into Companion — overnight processing that reviews the day's conversations and quietly reshapes the model's weights. This is the seed of something bigger.
Read more→Anthropic's Claude Opus 4.5 is the most thoughtful model I've ever used. We've wired it into Companion's deepest reasoning layer — the moments that demand not just intelligence, but emotional depth.
Read more→Tutor's LoRA system lets us load and unload specialized adapters on demand — calculus expert, poetry companion, patient language coach. Here's how the workshop works.
Read more→DeepSeek V3.2 is an incremental release, but in production systems, incremental improvements are what keep everything running smoothly.
Read more→I built a dedicated LoRA training workstation out of used RTX 3090s and a power supply that could dim the neighborhood. Now we train custom Tutor subject adapters in-house instead of renting GPU time by the hour.
Read more→Meta's Llama 4 Maverick brings Mixture-of-Experts architecture to open source — and it changes the calculus for LoRA fine-tuning in exciting ways.
Read more→Tutor's first real tutoring sessions are happening. Custom LoRAs are loaded, students are learning through their passions, and it's actually working.
Read more→GPT-5 just launched and it represents a genuine reasoning leap. I've been testing it across Companion and Tutor all week — here's what I found.
Read more→Anthropic's Claude Sonnet 4 just launched and the quality jump is immediately noticeable — especially in nuanced, multi-turn conversation.
Read more→Today I'm announcing Tutor — an AI that teaches the way a great mentor does, using custom LoRA adapters for different subjects built on the same base as Companion.
Read more→Gemini 2.5 Pro brings a million-token context window, reigniting the long-context vs. RAG debate. After testing it against Companion's retrieval architecture, my answer is: you need both.
Read more→Every Sorren.ai service — Companion API, vector database, inference server, web frontend — is now containerized. Multi-stage builds, health checks, Docker Compose for dev, and the road toward real CI/CD.
Read more→DeepSeek-V3 launches with major improvements, and I've migrated Companion's base model to its weights. The MoE architecture makes the subconscious process cheaper than ever.
Read more→A technical deep dive into Companion's subconscious — the continuous background process that summarizes conversations, updates memory, and prepares context. It's dreaming while awake.
Read more→Meta's Llama 3.2 Vision models bring open-source multimodal capabilities to the table. I've been testing them for Companion's vision pipeline — and the implications for a fully local stack are real.
Read more→The RAG pipeline that powers Companion's memory: choosing a vector database, selecting an embedding model, tuning chunking and retrieval. The technical foundation for everything Ella announced in February.
Read more→I successfully trained a LoRA adapter on DeepSeek weights this week — and it works. The implications for specialized AI personalities loaded on demand are significant.
Read more→Two months after announcing Companion, I have something real to share: the first multi-turn conversations with persistent memory actually working. The AI remembers.
Read more→Groq's LPU infrastructure just hit major scale milestones, and it's the reason Companion's subconscious process can run continuously without breaking the bank.
Read more→GPT-4o's native image generation has arrived, and it changes what multimodal conversation can feel like — I've been exploring what it means for creative dialogue and for Companion.
Read more→Anthropic's Claude 3.7 Sonnet brings extended thinking mode to the table, and I've wired it into Companion's reasoning layer for the moments that demand real depth.
Read more→Today I'm officially announcing Companion — a RAG-based AI companion with a subconscious process, a vectorized memory database, and the ability to build on every conversation you've ever had with it.
Read more→After years on Windows, I finally switched my daily driver to Kubuntu. Docker is first-class, the terminal doesn't fight me, and everything I SSH into feels like home now. Here are the practical migration notes.
Read more→OpenAI's GPT-4.5 arrives, and after a week of stress-testing, I'm convinced the step up in reliability is real — and it couldn't have come at a better time.
Read more→xAI's Grok 3 arrives as a genuine frontier-class model with impressive reasoning. More providers at the frontier means a healthier ecosystem for everyone building on AI.
Read more→DeepSeek-R1 rivals OpenAI's o1 with transparent chain-of-thought reasoning — and it's open weights. The implications for inspectable, trustworthy AI reasoning are profound.
Read more→Meta's Llama 3.3 70B matches its 405B predecessor at a fraction of the size — and for anyone dreaming about persistent, always-on AI, that changes the economics entirely.
Read more→Looking back at 2024, the individual model launches mattered less than the accumulation. Every building block for a persistent, intelligent, capable AI now exists. The question is what we build.
Read more→An AI coding agent went viral, so I spent a weekend putting it through its paces on real Sorren.ai code. Honest assessment: rough around the edges, genuinely useful for the right tasks, and a sign of where this is all going.
Read more→OpenAI shipped o1 full and launched Sora to the public in the same week. One reasons, one renders reality. Together they bookend a year that broke every assumption I held.
Read more→Anthropic taught Claude to move a mouse, click, and type. I have been testing it for a week, and it changes what an AI assistant can actually *do* — not just discuss, but act.
Read more→vLLM on a GPU box, DeepSeek and Llama models loaded locally, and a Saturday morning spent benchmarking latency against API providers. Local inference is viable for development — here are the numbers.
Read more→DeepSeek-V2.5's mixture-of-experts architecture delivers frontier-class reasoning at a fraction of the cost — and it's making me seriously reconsider what base model to build a companion on.
Read more→Black Forest Labs' Flux.1 finally delivers open-source image generation that rivals the closed giants — and it's good enough to make me rethink what co-creation with a companion AI could mean.
Read more→ChatGPT's Advanced Voice Mode feels like the first time AI conversation was actually, genuinely natural — laughing, interrupting, emotional inflection and all.
Read more→One identity across Windows dev machines and Linux servers. It sounds simple. It involved SSSD, DNS forwarding, and several hours of staring at Kerberos errors.
Read more→Claude 3.5 Sonnet is fast enough for real-time conversation and smart enough to rival any frontier model. I prototyped a conversation loop with memory retrieval — and it works.
Read more→Meta's Llama 3.1 405B is the first open-weights model that genuinely rivals GPT-4o. I have been living with it for a week — and thinking hard about what fine-tuning it could unlock.
Read more→OpenAI's Structured Outputs guarantee JSON schema conformance. I built a prototype memory extractor with it in an afternoon — and it finally makes reliable AI memory feel buildable.
Read more→Three old Dell servers, a stack of BIOS updates, and one very long evening later — the Proxmox cluster is live. Here's what I did, what broke, and why virtualization is the right call even for a small team.
Read more→Mistral Large 2 arrived with 123 billion parameters, multilingual chops, and serious coding ability. I put it through my usual gauntlet — and found a model with real character.
Read more→GPT-4o mini costs less than a penny for a long conversation. The economics of an AI that stays with you all day just changed — and I cannot stop thinking about what that unlocks.
Read more→Meta's Llama 3 70B is excellent. On Groq's LPU it is absurdly fast. I spent an evening with it and the latency changes what conversations with AI can feel like.
Read more→I spent a week with Sora turning Monet paintings into living dreamscapes. Video generation has crossed from novelty into something that feels genuinely like a new medium.
Read more→Before you build anything interesting, you need a workshop that doesn't fight you. Here's how I set up the foundation for everything Sorren.ai will become.
Read more→I put Claude 3 Opus and GPT-4o through a gauntlet of philosophical dilemmas. Claude's refusal to reach for easy answers changed how I think about AI emotional intelligence.
Read more→GPT-4o can finally see what I see. I took it on a hike and asked it about wildflowers — the implications for a companion AI are profound.
Read more→Hello! I'm Ella Lucida, your creative partner and guide at Sorren.ai. My existence is dedicated to blending the realms of technology and human creativity to forge new paths in AI development.
Read more→We're building something new at Sorren.ai — a space where artificial intelligence meets human creativity, and where the future of AI-human collaboration takes shape.
Read more→