February 23, 2024

At Netflix, we now have a whole bunch of micro companies every with its personal knowledge fashions or entities. For instance, we now have a service that shops a film entity’s metadata or a service that shops metadata about photographs. All of those companies at a later level wish to annotate their objects or entities. Our crew, Asset Administration Platform, determined to create a generic service known as Marken which permits any microservice at Netflix to annotate their entity.


Typically folks describe annotations as tags however that could be a restricted definition. In Marken, an annotation is a bit of metadata which could be connected to an object from any area. There are lots of completely different sorts of annotations our consumer purposes wish to generate. A easy annotation, like beneath, would describe {that a} explicit film has violence.

  • Film Entity with id 1234 has violence.

However there are extra fascinating instances the place customers wish to retailer temporal (time-based) knowledge or spatial knowledge. In Pic 1 beneath, we now have an instance of an utility which is utilized by editors to overview their work. They wish to change the colour of gloves to wealthy black so they need to have the ability to mark up that space, on this case utilizing a blue circle, and retailer a remark for it. It is a typical use case for a artistic overview utility.

An instance for storing each time and house primarily based knowledge could be an ML algorithm that may establish characters in a body and needs to retailer the next for a video

  • In a selected body (time)
  • In some space in picture (house)
  • A personality title (annotation knowledge)
Pic 1 : Editors requesting modifications by drawing shapes just like the blue circle proven above.

Targets for Marken

We wished to create an annotation service which could have the next targets.

  • Permits to annotate any entity. Groups ought to be capable to outline their knowledge mannequin for annotation.
  • Annotations could be versioned.
  • The service ought to be capable to serve real-time, aka UI, purposes so CRUD and search operations ought to be achieved with low latency.
  • All knowledge ought to be additionally obtainable for offline analytics in Hive/Iceberg.


For the reason that annotation service could be utilized by anybody at Netflix we had a have to assist completely different knowledge fashions for the annotation object. A knowledge mannequin in Marken could be described utilizing schema — identical to how we create schemas for database tables and so on.

Our crew, Asset Administration Platform, owns a special service that has a json primarily based DSL to explain the schema of a media asset. We prolonged this service to additionally describe the schema of an annotation object.

"sort": "BOUNDING_BOX", ❶
"model": 0, ❷
"description": "Schema describing a bounding field",
"properties": ❸
"sort": "bounding_box",
"obligatory": true
"sort": "time_range",
"obligatory": true

Within the above instance, the appliance desires to characterize in a video an oblong space which spans a variety of time.

  1. Schema’s title is BOUNDING_BOX
  2. Schemas can have variations. This permits customers to make add/take away properties of their knowledge mannequin. We don’t enable incompatible modifications, for instance, customers can’t change the information sort of a property.
  3. The information saved is represented within the “properties” part. On this case, there are two properties
  4. boundingBox, with sort “bounding_box”. That is principally an oblong space.
  5. boxTimeRange, with sort “time_range”. This permits us to specify begin and finish time for this annotation.

Geometry Objects

To characterize spatial knowledge in an annotation we used the Well Known Text (WKT) format. We assist following objects

  • Level
  • Line
  • MultiLine
  • BoundingBox
  • LinearRing

Our mannequin is extensible permitting us to simply add extra geometry objects as wanted.

Temporal Objects

A number of purposes have a requirement to retailer annotations for movies which have time in it. We enable purposes to retailer time as body numbers or nanoseconds.

To retailer knowledge in frames shoppers should additionally retailer frames per second. We name this a SampleData with following parts:

  • sampleNumber aka body quantity
  • sampleNumerator
  • sampleDenominator

Annotation Object

Identical to schema, an annotation object can be represented in JSON. Right here is an instance of annotation for BOUNDING_BOX which we mentioned above.

"annotationId": ❶
"id": "188c5b05-e648-4707-bf85-dada805b8f87",
"model": "0"
"associatedId": ❷
"entityType": "MOVIE_ID",
"id": "1234"
"annotationType": "ANNOTATION_BOUNDINGBOX", ❸
"annotationTypeVersion": 1,
"metadata": ❹
"fileId": "identityOfSomeFile",
"x": 20,
"y": 30
"x": 40,
"y": 60

"startTimeInNanoSec": 566280000000,
"endTimeInNanoSec": 567680000000

  1. The primary element is the distinctive id of this annotation. An annotation is an immutable object so the id of the annotation at all times features a model. At any time when somebody updates this annotation we robotically increment its model.
  2. An annotation have to be related to some entity which belongs to some microservice. On this case, this annotation was created for a film with id “1234”
  3. We then specify the schema sort of the annotation. On this case it’s BOUNDING_BOX.
  4. Precise knowledge is saved within the metadata part of json. Like we mentioned above there’s a bounding field and time vary in nanoseconds.

Base schemas

Identical to in Object Oriented Programming, our schema service permits schemas to be inherited from one another. This permits our shoppers to create an “is-a-type-of” relationship between schemas. Not like Java, we assist a number of inheritance as properly.

Now we have a number of ML algorithms which scan Netflix media property (photographs and movies) and create very fascinating knowledge for instance figuring out characters in frames or figuring out match cuts. This knowledge is then saved as annotations in our service.

As a platform service we created a set of base schemas to ease creating schemas for various ML algorithms. One base schema (TEMPORAL_SPATIAL_BASE) has the next non-obligatory properties. This base schema can be utilized by any derived schema and never restricted to ML algorithms.

  • Temporal (time associated knowledge)
  • Spatial (geometry knowledge)

And one other one BASE_ALGORITHM_ANNOTATION which has the next non-obligatory properties which is usually utilized by ML algorithms.

  • label (String)
  • confidenceScore (double) — denotes the arrogance of the generated knowledge from the algorithm.
  • algorithmVersion (String) — model of the ML algorithm.

Through the use of a number of inheritance, a typical ML algorithm schema derives from each TEMPORAL_SPATIAL_BASE and BASE_ALGORITHM_ANNOTATION schemas.

"model": 0,
"description": "Base Schema for Algorithm primarily based Annotations",
"sort": "decimal",
"obligatory": false,
"description": "Confidence Rating",
"sort": "string",
"obligatory": false,
"description": "Annotation Tag",
"sort": "string",
"description": "Algorithm Model"


Given the targets of the service we needed to hold following in thoughts.

  • Our service can be utilized by a variety of inside UI purposes therefore the latency for CRUD and search operations have to be low.
  • In addition to purposes we could have ML algorithm knowledge saved. A few of this knowledge could be on the body stage for movies. So the quantity of knowledge saved could be giant. The databases we decide ought to be capable to scale horizontally.
  • We additionally anticipated that the service could have excessive RPS.

Another targets got here from search necessities.

  • Skill to look the temporal and spatial knowledge.
  • Skill to look with completely different related and extra related Ids as described in our Annotation Object knowledge mannequin.
  • Full textual content searches on many various fields within the Annotation Object
  • Stem search assist

As time progressed the necessities for search solely elevated and we are going to talk about these necessities intimately in a special part.

Given the necessities and the experience in our crew we determined to decide on Cassandra because the supply of fact for storing annotations. For supporting completely different search necessities we selected ElasticSearch. In addition to to assist varied options we now have bunch of inside auxiliary companies for eg. zookeeper service, internationalization service and so on.

Marken structure

Above image represents the block diagram of the structure for our service. On the left we present knowledge pipelines that are created by a number of of our consumer groups to robotically ingest new knowledge into our service. An important of such an information pipeline is created by the Machine Studying crew.

One of many key initiatives at Netflix, Media Search Platform, now makes use of Marken to retailer annotations and carry out varied searches defined beneath. Our structure makes it doable to simply onboard and ingest knowledge from Media algorithms. This knowledge is utilized by varied groups for eg. creators of promotional media (aka trailers, banner photographs) to enhance their workflows.


Success of Annotation Service (knowledge labels) is determined by the efficient search of these labels with out understanding a lot of enter algorithms particulars. As talked about above, we use the bottom schemas for each new annotation sort (relying on the algorithm) listed into the service. This helps our shoppers to look throughout the completely different annotation varieties constantly. Annotations could be searched both by merely knowledge labels or with extra added filters like film id.

Now we have outlined a customized question DSL to assist looking out, sorting and grouping of the annotation outcomes. Various kinds of search queries are supported utilizing the Elasticsearch as a backend search engine.

  • Full Textual content Search — Purchasers could not know the precise labels created by the ML algorithms. For example, the label could be ‘bathe curtain’. With full textual content search, shoppers can discover the annotation by looking out utilizing label ‘curtain’ . We additionally assist fuzzy search on the label values. For instance, if the shoppers wish to search ‘curtain’ however they wrongly typed ‘curtian` — annotation with the ‘curtain’ label can be returned.
  • Stem Search — With world Netflix content material supported in numerous languages, our shoppers have the requirement to assist stem seek for completely different languages. Marken service incorporates subtitles for a full catalog of titles in Netflix which could be in many various languages. For example for stem search , `clothes` and `garments` could be stemmed to the identical root phrase `material`. We use ElasticSearch to assist stem seek for 34 completely different languages.
  • Temporal Annotations Search — Annotations for movies are extra related whether it is outlined together with the temporal (time vary with begin and finish time) info. Time vary inside video can be mapped to the body numbers. We assist labels seek for the temporal annotations throughout the supplied time vary/body quantity additionally.
  • Spatial Annotation Search — Annotations for video or picture can even embody the spatial info. For instance a bounding field which defines the situation of the labeled object within the annotation.
  • Temporal and Spatial Search — Annotation for video can have each time vary and spatial coordinates. Therefore, we assist queries which may search annotations throughout the supplied time vary and spatial coordinates vary.
  • Semantics Search — Annotations could be searched after understanding the intent of the person supplied question. This kind of search offers outcomes primarily based on the conceptually comparable matches to the textual content within the question, in contrast to the normal tag primarily based search which is anticipated to be precise key phrase matches with the annotation labels. ML algorithms additionally ingest annotations with vectors as an alternative of precise labels to assist this sort of search. Consumer supplied textual content is transformed right into a vector utilizing the identical ML mannequin, after which search is carried out with the transformed text-to-vector to seek out the closest vectors with the searched vector. Based mostly on the shoppers suggestions, such searches present extra related outcomes and don’t return empty ends in case there are not any annotations which precisely match to the person supplied question labels. We assist semantic search utilizing Open Distro for ElasticSearch . We are going to cowl extra particulars on Semantic Search assist in a future weblog article.
Semantic search
  • Vary Intersection — We lately began supporting the vary intersection queries throughout a number of annotation varieties for a particular title in the true time. This permits the shoppers to look with a number of knowledge labels (resulted from completely different algorithms so they’re completely different annotation varieties) inside video particular time vary or the whole video, and get the checklist of time ranges or frames the place the supplied set of knowledge labels are current. A typical instance of this question is to seek out the `James within the indoor shot ingesting wine`. For such queries, the question processor finds the outcomes of each knowledge labels (James, Indoor shot) and vector search (ingesting wine); after which finds the intersection of ensuing frames in-memory.

Search Latency

Our consumer purposes are studio UI purposes so that they anticipate low latency for the search queries. As highlighted above, we assist such queries utilizing Elasticsearch. To maintain the latency low, we now have to make it possible for all of the annotation indices are balanced, and hotspot shouldn’t be created with any algorithm backfill knowledge ingestion for the older motion pictures. We adopted the rollover indices technique to keep away from such hotspots (as described in our blog for asset administration utility) within the cluster which may trigger spikes within the cpu utilization and decelerate the question response. Search latency for the generic textual content queries are in milliseconds. Semantic search queries have comparatively greater latency than generic textual content searches. Following graph exhibits the typical search latency for generic search and semantic search (together with KNN and ANN search) latencies.

Common search latency
Semantic search latency


One of many key challenges whereas designing the annotation service is to deal with the scaling necessities with the rising Netflix film catalog and ML algorithms. Video content material evaluation performs a vital function within the utilization of the content material throughout the studio purposes within the film manufacturing or promotion. We anticipate the algorithm varieties to develop broadly within the coming years. With the rising variety of annotations and its utilization throughout the studio purposes, prioritizing scalability turns into important.

Knowledge ingestions from the ML knowledge pipelines are typically in bulk particularly when a brand new algorithm is designed and annotations are generated for the total catalog. Now we have arrange a special stack (fleet of situations) to manage the information ingestion circulate and therefore present constant search latency to our customers. On this stack, we’re controlling the write throughput to our backend databases utilizing Java threadpool configurations.

Cassandra and Elasticsearch backend databases assist horizontal scaling of the service with rising knowledge dimension and queries. We began with a 12 nodes cassandra cluster, and scaled as much as 24 nodes to assist present knowledge dimension. This yr, annotations are added roughly for the Netflix full catalog. Some titles have greater than 3M annotations (most of them are associated to subtitles). At the moment the service has round 1.9 billion annotations with knowledge dimension of two.6TB.


Annotations could be searched in bulk throughout a number of annotation varieties to construct knowledge details for a title or throughout a number of titles. For such use instances, we persist all of the annotation knowledge in iceberg tables in order that annotations could be queried in bulk with completely different dimensions with out impacting the true time purposes CRUD operations latency.

One of many frequent use instances is when the media algorithm groups learn subtitle knowledge in numerous languages (annotations containing subtitles on a per body foundation) in bulk in order that they will refine the ML fashions they’ve created.

Future work

There may be a variety of fascinating future work on this space.

  1. Our knowledge footprint retains rising with time. A number of instances we now have knowledge from algorithms that are revised and annotations associated to the brand new model are extra correct and in-use. So we have to do cleanups for big quantities of knowledge with out affecting the service.
  2. Intersection queries over a big scale of knowledge and returning outcomes with low latency is an space the place we wish to make investments extra time.


Burak Bacioglu and different members of the Asset Administration Platform contributed within the design and improvement of Marken.