Guides
Vendor-neutral, engineer-written guides to the concepts behind multimodal AI: perception, retrieval, embeddings, and the infrastructure agents use to see, hear, and search unstructured data. Learn the idea first; then see how Mixpeek applies it.
103 guides across 17 topics
How to Build a Taxonomy From Unlabeled Data: Clustering, Labeling, and Promotion
You have a million files and no categories. This walks the whole path: embedding the content, clustering it without guessing how many groups exist, naming what comes out, and the step almost everyone skips, which is deciding which clusters actually deserve to become categories. Covers HDBSCAN against k-means, why silhouette scores mislead on embeddings, exemplar-based labeling, promotion criteria, drift once new content arrives, and how the picture changes when the corpus is video or audio rather than text.
What Is Hybrid Search? BM25, Vector Retrieval, and How to Fuse Their Rankings
Why keyword and vector retrieval fail on opposite queries, what BM25 actually computes, and how Reciprocal Rank Fusion combines two rankings whose scores are not on the same scale. Covers the term-frequency saturation and length normalization inside BM25, why raw score addition breaks, RRF versus min-max and convex combination, how to pick the weighting, when hybrid is worse than either half, and what changes when one side of the index is video or images.
What Is MUVERA? Turning Multi-Vector Retrieval Into a Single-Vector Search
How MUVERA encodes a whole set of ColBERT-style token embeddings into one Fixed Dimensional Encoding whose inner product approximates the MaxSim score, so multi-vector retrieval can run on ordinary MIPS indexes. Covers the SimHash partitioning, the query-sum versus document-centroid asymmetry that makes the approximation hold, empty-cluster filling, how repetitions set the final dimension, and why the FDE is a candidate generator that still needs exact rescoring.
How Do You Run Your Own Model Inside a Managed Search Pipeline?
The four places a managed retrieval platform can let you run your own code (ingest-time extraction, query-time inference, reranking, and enrichment), what each one demands of the platform, and why the ingest and query sides must load the identical model or your vectors and your queries end up in different spaces. Covers the packaging contract, the cold-start and GPU-allocation problems that decide whether query-time custom inference is usable, versioning against an already-indexed corpus, and how to tell a real extension point from a webhook with good marketing.
How Do You Isolate Tenants in a Vector Index?
The three architectures for multi-tenant vector search (index per tenant, shared index with a tenant filter, and namespaces or partitions), what each costs, and the failure modes specific to each. Covers why a tenant filter is a correctness boundary rather than a performance knob, why post-filtering silently returns fewer results than you asked for, and why an unindexed tenant field can turn a working filter into an empty result set.
What Does It Cost to Make a Video Library Searchable?
A vendor-neutral cost model for turning raw video, images, and documents into a searchable index: the four cost centers (extraction, vector storage, serving, re-extraction), why per-feature extraction pricing stacks into the real bill, the arithmetic behind vector storage, and the re-extraction multiplier that decides whether you can ever upgrade your embedding model. With a worked 1,000-hour example using published July 2026 prices.
How to Search, Deduplicate, and Moderate AI-Generated Video and Images (FLUX 3, Kling, Veo, Runway)
AI video and image generators like FLUX 3 create content but do not search, deduplicate, or moderate the library you produce. A vendor-neutral guide to the retrieval layer that does: extraction (scenes, transcripts, faces, embeddings, fingerprints), semantic search, near-duplicate detection, pre-publication moderation, and clustering over your generated media in object storage.
Video RAG: Building Retrieval-Augmented Generation Over Video Corpora
A deep technical guide to building RAG pipelines over video data. Covers scene-level chunking, dual-channel indexing (visual + transcript), keyframe selection for LVLMs, temporal retrieval, and production architecture patterns.
How Does LoRA Fine-Tuning Work? (Adapters, QLoRA, DoRA, and Fine-Tuning Retrieval Models)
LoRA fine-tunes a large model by freezing its weights and training a tiny low-rank adapter (a few hundred MB, under 1% of parameters): trainable on one GPU, swappable at inference. How LoRA, QLoRA, and DoRA work; the rank/alpha/target-module knobs that matter in 2026; and the retrieval-specific part most guides skip: fine-tuning embedding models, rerankers, and VLMs with LoRA, why you must re-embed your corpus afterward, and how swappable adapters enable multi-tenant retrieval. Vendor-neutral, with a comparison table and FAQs.
How Do I Evaluate a RAG Pipeline? (Faithfulness, Answer Relevance, Hallucination, and Context Metrics)
Evaluating RAG means scoring two different things separately: did retrieval fetch the right context (context precision and recall), and did generation use it honestly (faithfulness/groundedness and answer relevance). The four metrics that matter, how LLM-as-judge actually scores faithfulness and where it goes wrong, how to build a golden eval set, why you must measure retrieval and generation independently to know which half is broken, the multimodal wrinkle, and how to run it on a Mixpeek pipeline. Vendor-neutral, with the RAGAS-style metric definitions.
How Do I Ingest Millions of Files into a Search Index? (Images, Video, Documents at Scale)
Ingesting millions of files is not a bigger version of ingesting a hundred: it is a different problem governed by five constraints, batching, bounded concurrency, backpressure, resumability, and idempotent retries that don't re-pay for work already done. The scaling architecture (chunk into batches, cap in-flight work, checkpoint so a failure resumes instead of restarting, and dedup so a retry reuses prior extraction instead of re-running the GPU), the cost trap that quietly doubles GPU spend, how it differs from a one-off script, and how to run it on Mixpeek.
How Do AI Agents Search Big Datasets by Navigating Clusters? (Hierarchical Cluster Search)
Flat vector search returns top-k against one query vector, which breaks down when an agent does not know the right query, the corpus is huge and diverse, or the task is exploratory. Agentic hierarchical cluster search gives the agent a map instead: a cluster hierarchy (themes -> sub-clusters -> records) it navigates coarse-to-fine, scoring its goal against a few dozen centroids and drilling into the matching branch before running a precise retrieval at the leaf. When it beats flat ANN, the navigation loop, the cost math, the honest limits, and how to build it from clustering + composite clustering + a cluster-scoped retriever.
From concept to production
These guides explain how multimodal perception and retrieval actually work. Mixpeek is the platform that runs them: point it at your storage and get back relevant, timestamped results.