Greater than 20k clients – together with over 60% of the Fortune 500 – depend on Databricks to detect fraud, uncover new medication, optimize their provide chain, personalize buyer experiences, and extra. And as clients ask us to broaden our product, they’re additionally asking us to broaden who we companion with to assist them democratize information and AI throughout their enterprise… Irrespective of how briskly we innovate, we’ll by no means be capable to meet each buyer’s wants alone.
That is the place you – our ISV and Information Supplier companions – are available. Collectively, we aspire to supply clients any instrument, mannequin, dataset, or agent they want, and assist develop your corporation alongside the way in which.
If you wish to go quick, go alone. If you wish to go far, go collectively. —African Proverb
Within the 10 months since becoming a member of Databricks to take care of our Product Ecosystem, I’ve had the privilege of assembly a whole bunch of ISVs, Information Suppliers, and startups. You advised me what’s working properly and the place we have to enhance, and we have lots in retailer to handle your suggestions.
At this time at our Accomplice Kickoff in Vegas, I am excited to announce the primary two (of many) enhancements to our ISV and Information Supplier companion program: a brand new Accomplice Tiering Program to raised reward companions for his or her buyer impression throughout a number of companion architectures, and a Accomplice Properly-Architected Framework that offers clear, AI-ready technical steering for constructing world-class integrations with Databricks.
Three Accomplice Architectures for ISVs and Information Suppliers on Databricks
On the basis of those bulletins are the three architectures you co-build with us.
1. Linked Merchandise are partner-built merchandise that combine with the shopper’s Databricks platform to investigate, course of, rework, and generate insights from their information. These drivers (e.g., JDBC, ODBC, Python) and APIs are mostly developed by AI, BI, information governance, information transformation, or different data-intensive instruments, and now additionally consists of MCP servers that may be printed to our MCP Market.
2. Delta Sharing is the one “one share to anyplace protocol” that permits Information Suppliers to share their information with clients on any platform, utilizing any instrument, and energy any mannequin. It’s zero-copy, so that you don’t have to duplicate your information, and we handle all of the cross-cloud networking and safety complexity. Not like the “walled backyard” approaches supplied by different information platforms, we’ll proceed to make Delta Shares work with any instrument clients need. If the recipient accepts the share in Databricks, it merely seems of their Unity Catalog to make use of/be part of/question alongside the remainder of their information, and it will additionally work in Excel, any main BI instrument, Snowflake, and others. That is probably why Delta Shares have grown greater than 140% throughout the ecosystem this 12 months.
3. Constructed-Ons are options constructed on high of Databricks, with Databricks serving because the foundational implementation behind the companion’s personal front-end/API/mental property. That is the latest movement for us, and the one which we’re seeing probably the most demand for, significantly as ISVs construct brokers and functions with Agent Bricks, Lakebase, and Databricks Apps.
Beforehand, we have had totally different groups with totally different incentives pursuing partnerships throughout these architectures. This made sense as every new structure was launched, as we needed to take a look at, measure, and be taught what labored. What we realized was that a lot of you interact in a couple of of those architectures and should be rewarded accordingly, and that you really want higher steering on *how* to construct your integrations to satisfy buyer expectations and get help from our area.
One Program, Three Accomplice Architectures
Our new Accomplice Tiering Program makes use of the identical tiering construction as our Consulting & System Integrator program, and measures the mixed impression of all three companion architectures – variety of joint clients, Databricks consumption ($DBU) impression, dedication to Databricks, strategic providers adoption, and GTM readiness – to find out tier placement and rewards in a holistic, merit-based system.
What this implies for you:
- Get credit score for every thing you construct throughout Linked Merchandise, Delta Sharing, and Constructed-Ons
- Get “additional” credit score for constructing with Lakebase, Genie, and/or Agent Bricks.
- See how your structure decisions translate into tiers, advantages, and GTM help
We’ll be sharing extra with you in regards to the particular tiers and advantages all through Q1(Feb-Apr) and proceed to refine based mostly in your suggestions earlier than we formally rollout on or earlier than Q2.
Accomplice Properly Architected Framework: Constructing Higher, Quicker, Collectively
To offer you higher technical steering and higher consistency/transparency in how we validate all companion integrations, we’re publishing the Accomplice Properly-Architected Framework (PWAF).
PWAF supplies companions with AI-ready prescriptive steering for constructing integrations which might be safe and dependable, and consists of the instrumentation necessities that permit us to measure buyer adoption and $DBU impression – a number of the key metrics that decide your tier placement within the Accomplice Program.
Better of all, we have designed PWAF to be AI-ready from the bottom up. In just some clicks, you should utilize this steering with Cursor, Claude Code, Replit, or your favourite AI-assisted IDE/developer instrument and immediate your technique to well-architected Databricks integrations.
And, fairly than simply inform you what to do, we need to present you the way. To accompany PWAF, we’ve constructed a fictitious firm “Firefly Analytics” that’s been architected to observe PWAF’s greatest practices and provide the scaffolding to get began from a sensible instance. 100% vibe coded, and (quickly) 100% open supply.
Level your instruments at PWAF and the Firefly repo coming quickly and get constructing!
And we’re simply getting began
These are the primary of many enhancements we’re engaged on that can assist you construct transformative capabilities and develop your corporation by delighting our joint clients.
Thanks for all of your suggestions to assist us construct a greater program, and we’re wanting ahead to rising and constructing collectively this 12 months!
Hold constructing,
—Stephen
