{"id":51320566,"url":"https://github.com/redis/redis-ai-research-public","last_synced_at":"2026-07-01T13:00:51.042Z","repository":{"id":368469684,"uuid":"1224437952","full_name":"redis/redis-ai-research-public","owner":"redis","description":"Home of AI Research at Redis","archived":false,"fork":false,"pushed_at":"2026-06-30T17:10:20.000Z","size":1799,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-30T19:09:41.028Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/redis.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-29T09:28:22.000Z","updated_at":"2026-06-30T17:10:39.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/redis/redis-ai-research-public","commit_stats":null,"previous_names":["redis/redis-ai-research-public"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/redis/redis-ai-research-public","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/redis%2Fredis-ai-research-public","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/redis%2Fredis-ai-research-public/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/redis%2Fredis-ai-research-public/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/redis%2Fredis-ai-research-public/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/redis","download_url":"https://codeload.github.com/redis/redis-ai-research-public/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/redis%2Fredis-ai-research-public/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35007278,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-01T02:00:05.325Z","response_time":130,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-07-01T13:00:50.402Z","updated_at":"2026-07-01T13:00:51.036Z","avatar_url":"https://github.com/redis.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Redis AI Research\n\nPublic-facing projects, experiments, and reference implementations from the **Redis AI Research team**.\n\nThis monorepo collects open-source work exploring how Redis can be applied to modern AI workloads — semantic caching, vector search, retrieval-augmented generation (RAG), agentic systems, and beyond.\n\n## What you'll find here\n\nEach top-level directory is a self-contained project with its own README, dependencies, and setup instructions. Projects generally fall into one of these areas:\n\n- **Semantic \u0026 vector search** — using Redis as a vector store for embeddings, similarity search, and hybrid retrieval.\n- **Caching for LLM applications** — semantic caching, result reuse, and latency optimization patterns.\n- **Agentic systems** — multi-agent orchestration, learning agents, and memory architectures backed by Redis.\n- **RAG pipelines** — reference implementations and benchmarks for retrieval-augmented generation.\n- **Benchmarks \u0026 evaluations** — performance studies and methodology for AI workloads on Redis.\n\n## Repository layout\n\n```\n.\n├── coding-agents/              # multi-tenant OpenAI Agents SDK coding-agent\n│                               # service with sandboxed tenant workspaces\n├── learning-agents/            # multi-agent NL → pandas analysis with Redis-backed\n│                               # semantic cache and learned-guidance memory\n├── memorydb-semantic-search/   # vector search over ~238k LongMemEval sessions on\n│                               # AWS MemoryDB (pure-Redis vs. hybrid+S3 vs. hybrid+EBS)\n├── opencode-spec-optimization/ # iteratively rewrites AGENTS.md until OpenCode\n│                               # runs meet per-objective success thresholds\n└── ...\n```\n\nProjects are added incrementally. Browse the top-level directories to see what's currently available.\n\n### Current projects\n\n- **[coding-agents](coding-agents/)** — A multi-tenant FastAPI coding-agent service built on the OpenAI Agents SDK sandbox harness. Authenticates tenants, maps each request to an isolated workspace, supports synchronous and SSE-streamed task execution, and can fan work out to parallel sandbox subagents.\n- **[learning-agents](learning-agents/)** — A multi-agent system that turns natural-language questions into executable pandas code over a CSV/JSON dataset. Uses Redis vector indices for a semantic result cache and a persistent \"guidance\" memory that learns from past errors to improve future retries.\n- **[memorydb-semantic-search](memorydb-semantic-search/)** — Semantic search over ~238k LongMemEval chat sessions on AWS MemoryDB (Valkey 7.2 with search). Benchmarks three backend approaches — pure Redis, hybrid+S3, and hybrid+EBS — with Terraform for the full AWS stack (MemoryDB cluster, bastion, S3, VPC endpoint) and a FastAPI server exposing the search APIs.\n- **[opencode-spec-optimization](opencode-spec-optimization/)** — Workflow that runs every question in `QUESTIONS.yaml` against [OpenCode](https://opencode.ai) multiple times in parallel, evaluates the responses against the thresholds in `OBJECTIVE.yaml`, and iteratively rewrites `AGENTS.md` until every objective passes or the retry limit is hit. Useful for tuning agent system prompts via measurable, repeatable evals.\n\n## Getting started\n\n1. Clone the repository:\n   ```bash\n   git clone https://github.com/redis/redis-ai-research-public.git\n   cd redis-ai-research-public\n   ```\n2. Pick a project directory and follow its own README for setup and usage. Most projects assume a running Redis instance (Redis Stack or Redis 8+ with the search/vector modules).\n\n## Prerequisites (common)\n\n- Redis Stack or Redis 8+ (for vector search and JSON support)\n- Python 3.11+ (most projects)\n- An API key for the LLM provider used by the project (typically OpenAI, Anthropic, or Google)\n\nCheck each project's README for exact requirements.\n\n## Contributing\n\nContributions, issues, and discussion are welcome. Open an issue to propose a new experiment, report a bug, or ask a question. Pull requests should target the relevant project directory and include updates to that project's README where appropriate.\n\n## License\n\nThis repository is released under the [MIT License](LICENSE). See the `LICENSE` file for details.\n\n## Contact\n\nFor questions about the AI Research team's work, open an issue in this repository.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fredis%2Fredis-ai-research-public","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fredis%2Fredis-ai-research-public","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fredis%2Fredis-ai-research-public/lists"}