Alma by Olivares AI

Alma by Olivares AI is a comprehensive, memory-first artificial intelligence assistant designed to solve the persistent frustration of context loss that plagues standard AI chatbots. Unlike conventional tools that treat each conversation as a fresh start, Alma automatically extracts and retains facts, preferences, decisions, and behavioral patterns from every interaction it has with you. This creates a continuously evolving, long-term understanding of who you are and what you need across all sessions. The core value proposition is a unified workspace that integrates persistent memory, background agents, creative studios for image, video, and music generation, and deep research tools under a single, predictable monthly budget starting at just $14 per month with no hidden add-ons. Alma is built for a wide audience ranging from casual users who want a smarter conversational partner that remembers their preferred tone and ongoing projects, to creators producing thumbnails, soundtracks, and short videos without juggling separate subscriptions. It also serves researchers tracking trends and conducting deep web analysis that consolidates findings directly into memory, and power users who want to bring their own API keys or connect Alma to their editor via the Model Context Protocol (MCP). The platform supports 16 languages, offers voice conversations and voice note transcription, and provides a long-form editor with export capabilities to DOCX, XLSX, PPTX, and PDF formats. With its Soul Engine v2 cognitive framework, Alma goes beyond simple prompt engineering to define identity, worldview, style guides, and communication modes for a truly personalized AI experience.

Persistent Memory System

Alma features a complete memory system that automatically extracts facts, preferences, and decisions from every conversation. This is not a simple vector store but a contextual engine that understands patterns, personality, and learned behavior. Your AI remembers your preferred writing style, ongoing projects, and specific details shared weeks ago, ensuring continuity across all interactions without manual prompt engineering.

Background Agents

Users can describe a goal in plain language, and Alma plans and executes the task in the background for hours or even weeks. Agents can be configured to investigate competitors for a set period, watch the web for brand mentions, or generate recurring status reports from recent memories. They report back with scheduled briefs and final summaries, providing autonomous research and monitoring capabilities.

Unified Creative Studios

Three creative studios for image, video, and music generation are bundled into one workspace under a single monthly budget. The Image Studio uses Flux Pro and Leonardo models for generating assets across styles and sizes with inpainting and upscaling. The Video Studio leverages Runway Gen-4 and Gen-4.5 for storyboarding and stitching up to 60 seconds of 1080p video. The Music Studio uses ElevenLabs for studio-grade tracks up to 10 minutes, including focus loops and full demos with or without vocals.

Soul Engine v2 Cognitive Framework

Soul Engine v2 allows users to define a complete cognitive framework for their AI, including identity, worldview, style guide, anti-patterns, and communication modes. This goes far beyond a simple personality prompt, enabling the AI to understand context, personality, and learned behavior. It automatically assembles the right context for each interaction without requiring manual prompt engineering, ensuring consistent and appropriate responses.

Personalized Conversational Assistant for Professionals

Professionals can use Alma as a daily assistant that remembers the context of ongoing projects, preferred communication styles, and specific decisions made in previous conversations. For example, a solo founder working on a persistent memory SaaS can ask for concise, direct answers without filler and receive Spanish responses when writing copy. Alma maintains this knowledge across all sessions, eliminating the need to repeat context and increasing productivity.

Content Creation for Digital Creators

Creators can generate thumbnails, soundtracks, and short videos for social media or marketing campaigns all within a single workspace. Using the Image Studio, they can create and upscale assets with Flux Pro models. The Video Studio allows for storyboarding and generating up to 60 seconds of 1080p video. The Music Studio provides full tracks up to 10 minutes for intros, demos, or background loops. All assets are drawn from the same monthly budget without separate subscriptions.

Deep Research and Trend Monitoring

Researchers and knowledge workers can set background agents to investigate competitors or watch for brand mentions over extended periods. Alma conducts deep web search, summarizes findings directly into its persistent memory, and tracks trends for ongoing topics. The long-form editor supports exporting to PDF, DOCX, and Markdown, while keeping research threads separate across multiple projects to avoid contaminating context.

Development Workflow Integration for Power Users

Software developers and power users can bring their own provider keys (BYOK) for compatible AI models, connect Alma to their editor over the Model Context Protocol (MCP), or hit the API directly. The Max plan includes a VS Code extension with sidebar chat, allowing developers to interact with their persistent memory and background agents directly within their development environment for seamless workflow integration.

How does Alma's persistent memory work across different conversations?

Alma automatically extracts facts, preferences, and decisions from every interaction and stores them in a continuously evolving memory system. When you start a new conversation, the AI automatically assembles the right context from your memory, including your preferred writing style, ongoing projects, and specific details you shared previously. This eliminates the need to re-introduce yourself or repeat information, as the AI remembers everything across all sessions.

Can I use Alma for generating images, videos, and music without separate subscriptions?

Yes, Alma bundles all three creative studios under a single monthly budget with no add-ons or separate subscriptions. The Image Studio uses Flux Pro and Leonardo models, the Video Studio leverages Runway Gen-4 and Gen-4.5, and the Music Studio uses ElevenLabs for studio-grade audio. All generation is drawn from the same USD pool, and you can top up only when you exhaust your monthly budget.

What is Soul Engine v2 and how is it different from a standard personality prompt?

Soul Engine v2 is a complete cognitive framework that allows you to define your AI's identity, worldview, style guide, anti-patterns, and communication modes. Unlike a standard personality prompt which is a simple text instruction, Soul Engine v2 creates a structured understanding of how the AI should think and respond. It automatically assembles the right context for each interaction without manual prompt engineering, ensuring consistent behavior based on your defined parameters.

What export formats are supported for the long-form editor?

The long-form editor supports export to multiple document formats including DOCX for Microsoft Word, XLSX for Excel spreadsheets, PPTX for PowerPoint presentations, and PDF for portable document format. This allows users to generate and export long-form content, reports, or research findings directly from Alma into standard office applications for further editing or sharing.