Home Venture Capital The 2023 MAD (Machine Studying, Synthetic Intelligence & Information) Panorama – Matt Turck

The 2023 MAD (Machine Studying, Synthetic Intelligence & Information) Panorama – Matt Turck

The 2023 MAD (Machine Studying, Synthetic Intelligence & Information) Panorama – Matt Turck


It has been lower than 18 months since we revealed our final MAD panorama, and it has been filled with drama.

After we left, the information world was booming within the wake of the big Snowflake IPO, with an entire ecosystem of startups organizing round it. 

Since then, after all, public markets crashed, a recessionary economic system appeared and VC funding dried up. An entire era of information/AI startups has needed to adapt to a brand new actuality.

In the meantime, the previous few months noticed the unmistakable, exponential acceleration of Generative AI, with arguably the formation of a brand new mini-bubble. Past technological progress, it feels that AI has gone mainstream, with a broad group of non-technical folks around the globe now attending to expertise its energy firsthand.

The rise of information, ML and AI is likely one of the most basic developments in our era. Its significance goes effectively past the purely technical, with a deep influence on society, politics, geopolitics and ethics.

But it’s a sophisticated, technical, and quickly evolving world that’s typically complicated even for practitioners within the area. There’s a jungle of acronyms, applied sciences, merchandise and firms on the market which are onerous to maintain observe of, not to mention grasp:

The annual MAD (Machine Studying, Synthetic Intelligence and Information) panorama is our try at making sense of this vibrant area.  Its basic philosophy, very like our occasion collection Information Pushed NYC, has been to open supply work that we’d do anyway, and begin a dialog with the neighborhood.

So, right here we’re once more, in 2023. That is our ninth annual panorama and “state of the union” of the information and AI ecosystem. Listed here are the prior variations: 2012, 2014, 2016, 2017, 2018, 2019 (Half I and Half II), 2020 and 2021

This annual state of the union put up is organized in 4 elements:

After a lot analysis and energy, we’re proud to current the 2023 model of the MAD panorama. Once I say “we”, I imply a little bit group, whose nights can be haunted for months to return by recollections of shifting tiny logos out and in of crowded little bins on a PDF: Katie Mills, Kevin Zhang and Paolo Campos. Immense because of them. And sure, I meant it once I informed them on the onset “oh, it’s a light-weight venture, perhaps a day or two, it’ll be enjoyable, please signal right here”.

So, right here it’s (cue in drum roll, smoke machine).  The MAD panorama is available in two modes of consumption this yr:

PDF (static) model:

<<<<<<<< CLICK HERE FOR PDF VERSION >>>>>>>>

(sure, it’s all very excessive decision, and you may simply zoom on each desktop and cell)

<New!> Interactive model:

As well as, this yr for the primary time, we’re leaping head first into what the children name the “World Extensive Internet”, with a completely interactive model of the MAD Panorama that ought to make it enjoyable to discover the varied classes.  


Notes on the interactive model:

  • Every brand is clickable – once you click on a pop up reveals up on the underside proper nook
  • There’s a “panorama” and a “card” view (see high proper nook)… and in addition, an evening mode!
  • This can be a first model, and we’ll add extra performance ASAP (search, filtering, and many others.)
  • For this interactive model, we partnered with Gotta Go Quick for the app construct and CB Insights for the information that seems within the playing cards.  Many because of each for his or her partnership. 

For all questions and feedback, please e mail MAD2023@firstmarkcap.com 

Basic strategy

First, we’ve made the choice this yr once more to hold each knowledge infrastructure and ML/AI on the identical panorama. One may argue that these two worlds are more and more distinct. Nevertheless, we proceed to consider that there’s an important symbiotic relationship between these areas. Information feeds ML/AI fashions. The excellence between a knowledge engineer and a machine studying engineer is usually fairly fluid. Enterprises have to have a strong knowledge infrastructure in place so as earlier than correctly leveraging ML/AI.

The panorama is constructed roughly on the identical construction as each annual panorama since our first model in 2012. The free logic is to comply with the move of information, from left to proper – from storing and processing to analyzing to feeding ML/AI fashions and constructing user-facing, AI-driven or data-driven purposes.

This yr once more, we’ve stored a separate “open supply” part. It’s at all times been a little bit of an ungainly group as we successfully separate business firms from the open supply venture they’re typically the primary sponsor of. However equally, we need to seize the fact that for one open supply venture (for instance, Kafka), you’ve gotten many business firms and/or distributions (for Kafka – Confluent, Amazon, Aiven, and many others.). Additionally, some open supply initiatives showing within the field are usually not totally business firms but.

The overwhelming majority of the organizations showing on the MAD panorama are distinctive firms, with a really giant variety of VC-backed startups. Various others are merchandise (corresponding to merchandise provided by cloud distributors) or open supply initiatives.

Firm choice

This yr, now we have a complete of 1,416 logos showing on the panorama.   For comparability, there have been 139 in our first model in 2012.

Annually we are saying we will’t probably match extra firms on the panorama and every year, in some way, we have to. This comes with the territory of protecting some of the explosive areas of know-how.

Nevertheless, this yr particularly, we’ve needed to take a extra editorial, opinionated strategy to deciding which firms make it to the panorama. Regardless of the surging variety of firms within the class, we’re gone the stage the place we will match almost everybody, so now we have needed to make selections.

In prior years, we tended to offer disproportionate illustration to growth-stage firms, based mostly on funding stage (sometimes Sequence B-C or later) and ARR (when obtainable), along with all the massive incumbents. Nevertheless this yr, notably given the explosion of name new areas like Generative AI the place most firms are 1 or 2 years previous, we’ve made the editorial choice to function many extra very younger startups on the panorama.

A few disclaimers:

  • We’re VCs, so now we have a bias in direction of startups, though hopefully we’ve accomplished a superb job protecting bigger firms, cloud vendor choices, open supply and occasional bootstrapped firms
  • We’re based mostly within the US, so we most likely over-emphasize US startups. We do have sturdy illustration of European and Israeli startups on the MAD panorama. Nevertheless, whereas now we have a couple of Chinese language firms, we most likely under-emphasize the Asian market in addition to Latin America and Africa (which simply had a powerful knowledge/AI startup success with the acquisition of Tunisia-born Instadeep by BioNTech for $650M)


One of many tougher elements of the method is categorization – particularly, what to do when an organization’s product providing straddles two or extra areas. It’s changing into a extra salient subject yearly, as many startups progressively increase their providing, a pattern we focus on in “Half III – Information Infrastructure”.

Equally, it could be simply untenable to place each startup in a number of bins on this already overcrowded panorama.

Subsequently, our basic strategy has been to categorize an organization based mostly on its core providing, or what it’s largely recognized for.  Because of this, startups usually seem in just one field, even when they do greater than only one factor.

We make exceptions for the cloud hyperscalers (many AWS, Azure and GCP merchandise throughout the varied bins), in addition to some public firms (e.g. Datadog) or very giant personal firms (e.g., Databricks).

What’s new this yr

Most important adjustments in “Infrastructure”:

  • We (lastly) killed the Hadoop field, to replicate the gradual disappearance of the OG Huge Information know-how – the tip of an period! We had determined to maintain it one final time within the MAD 2021 panorama to replicate the present footprint. Hadoop is definitely not lifeless, and elements of the Hadoop ecosystem are nonetheless being actively used (e.g., Hive) – see The Hadoop Dialog Is Now About What’s Subsequent . Nevertheless it has declined sufficient that we determined to merge the varied distributors and merchandise supporting Hadoop into Information Lakes (and stored Hadoop and different associated initiatives in our Open Supply class).
  • Talking of information lakes, we rebranded that field to “Information Lakes / Lakehouses” to replicate the lakehouse pattern (which we had mentioned within the 2021 MAD panorama)
  • Within the ever evolving world of databases, we created three new subcategories:
    • “GPU-accelerated Databases” (used for streaming knowledge and real-time machine studying)
    • “Vector Databases” (used for unstructured knowledge to energy AI purposes, see What’s a Vector Database?)
    • “Database Abstraction”, a considerably amorphous time period meant to seize the emergence of a brand new group of serverless databases that summary away lots of the complexity concerned in managing and configuring a database. For extra, right here’s a superb overview: 2023 State of Databases for Serverless & Edge (mentions numerous distributors, greater than we may match within the field)
  • We thought-about including an Embedded Database” class with DuckDB for OLAP, KuzuDB for Graph, SQLite for RDBMS and Chroma for search however needed to make onerous selections given restricted actual property – perhaps subsequent yr.
  • We added a “Information Orchestration” field to replicate that rise of a number of business distributors in that area (we already had a “Information Orchestration” field in “Open Supply” in MAD 2021)
  • We merged two subcategories “Information observability” and “Information High quality” into only one field, to replicate the truth that firms within the area, whereas typically coming from completely different angles, are more and more overlapping – a sign that the class could also be ripe for consolidation.
  • We created a new “Totally Managed” knowledge infrastructure subcategory. This displays the emergence of startups that summary away the complexity of sewing collectively a series of information merchandise (see our ideas on the Fashionable Information Stack in Half III), saving their prospects time, not simply on the technical entrance, but in addition on contract negotiation, funds, and many others.

Most important adjustments in “Analytics”:

  • For now, we killed the “Metrics Retailer” subcategory we had created within the 2021 MAD panorama. The concept was that there was a lacking piece within the fashionable knowledge stack. The necessity for the performance definitely stays, nevertheless it’s unclear whether or not there’s sufficient there for a separate subcategory.  Early entrants within the area quickly developed: Supergrain pivoted, Hint* constructed an entire layer of analytics on high of its metrics retailer, and Remodel was just lately acquired by dbt Labs. 
  • We created a “Buyer Information Platform” field, as this subcategory, lengthy within the making, has been heating up.
  • On the danger of being “very 2022”, we created a “Crypto/web3 Analytics” field — we proceed to consider there are alternatives to construct necessary firms within the area.

Most important adjustments in “Machine Studying / Synthetic Intelligence”:

  • In our 2021 MAD panorama, we had damaged down “MLOps” into a number of subcategories – “Mannequin Constructing”, “Characteristic Shops” and “Deployment and Manufacturing”. On this yr’s MAD, we’ve merged all the pieces again into one large MLOps field. This displays the fact that many distributors’ choices within the area are actually considerably overlapping – one other class that’s ripe for consolidation.
  • We virtually created a brand new “LLMOps” class subsequent to MLOps to replicate the emergence of a brand new group of startups targeted on the particular infrastructure wants for giant language fashions. However the variety of firms there (not less than that we’re conscious of) remains to be too small and people firms actually simply bought began. 
  • We renamed “Horizontal AI” to “Horizontal AI / AGI” to replicate the emergence of an entire new group of research-oriented outfits, a lot of which overtly state Synthetic Basic Intelligence as their final aim.
  • We created a “Closed Supply Fashions” field, to replicate the unmistakable explosion of latest fashions over the past yr, particularly within the subject of Generative AI. We’ve additionally added a brand new field in “Open Supply” to seize the open supply fashions.
  • We added an “Edge AI” class – not a brand new subject, however there appears to be acceleration within the area

Most important adjustments in “Functions”:

  • We created a brand new “Functions/Horizontal” class, with subcategories corresponding to code, textual content, picture, video, and many others. The brand new field captures the explosion of latest Generative AI startups over the previous few months. After all, a lot of these firms are thin-layers on high of GPT and should or will not be round within the subsequent few years, however we consider it’s a essentially new necessary class and needed to replicate it on the 2023 MAD panorama. Observe that there are a couple of Generative AI startups talked about in “Functions/Enterprise” as effectively.
  • So as to make room for this new class:
    • We deleted the “Safety” field in “Functions/Enterprise”. We made this editorial choice as a result of, at this level, nearly each one of many 1000’s of safety startups on the market use ML/AI, and we may commit a complete panorama to them.
    • We trimmed down the “Functions/Trade” field. Particularly, as many bigger firms in areas like finance, well being or industrial have constructed some stage ML/AI into their product providing, we’ve made the editorial choice to focus totally on “AI-first” firms in these areas.

Different noteworthy adjustments:

  • We added a brand new ESG knowledge subcategory to “Information Sources & APIs” on the backside, to replicate its rising (if typically controversial) significance.

We significantly expanded our “Information Companies” class and rebranded it “Information & AI Consulting”, to replicate the rising significance of consulting providers to assist prospects dealing with a posh ecosystem, in addition to the truth that some pure-play consulting outlets are beginning to attain early scale.




Please enter your comment!
Please enter your name here