At Databricks, our automation imaginative and prescient is to automate all facets of the enterprise, making it higher, sooner, and cheaper. For the gross sales groups, we’re digitally reworking our vendor expertise by offering genAI brokers that help the vendor throughout the gross sales lifecycle. Our purpose is to enhance the vendor expertise with AI capabilities by seamlessly integrating them into their day-to-day duties and offering a less complicated, simpler approach for sellers to retrieve data property in addition to orchestrate actions by automating repetitive guide administrative duties.
Our “Subject AI Assistant” is constructed on the Databricks Mosaic AI agentic framework and supplies a approach for sellers to question and work together with information throughout a number of information sources. It integrates with a number of key platforms together with:
- Our Inside Databricks Lakehouse for account intelligence, gross sales enablement content material, and gross sales playbooks
- Our Buyer Relationship Administration Platform (CRM) system
- Our Collaboration Platform collates and indexes most of our nonstructured information
The AI utility is used to:
- Conversationally work together with information throughout a number of information sources utilizing pure language (beginning with English)
- Capability to obtain and create paperwork based mostly on the data gathered
- Take actions based mostly on the information insights (replace fields in our CRM, draft a customized outbound prospecting electronic mail, create a tailored buyer proposal, prep for a buyer assembly, and so on.
The sphere assistant responds to seeded prompts based mostly on person and web page context and likewise supplies a chat-like interface for open-ended queries on the above-mentioned datasets.
Enterprise Impression
Sellers are usually overwhelmed with the quantity of knowledge thrown at them. They want entry to information residing in numerous siloed functions, as a part of their regular day-to-day routine. They require quick access to account, alternative, and use case information that resides in our CRM, in addition to buyer market insights and account intelligence, together with account consumption information that resides in our lakehouse. As well as, additionally they want entry to gross sales content material – enablement playbooks, aggressive gross sales collateral in addition to product data base articles and product roadmap paperwork. It isn’t simply restricted to information retrieval, however the true effectivity positive factors happen when the repetitive guide duties they carry out each day based mostly on the information insights they retrieve could be totally automated. That’s precisely what the function of the sector AI assistant is – assist the sellers within the day-to-day duties together with data retrieval, distilling the insights from the data, and performing actions based mostly on these insights.
Answer Overview
Utilizing the Databricks Mosaic AI agent framework, we constructed a discipline AI assistant by integrating each structured and unstructured information from a number of information sources. The answer supplies a complete strategy personalised and tailor-made for our sellers, out there on-demand in our CRM. A few of the capabilities supplied embody:
Buyer insights present a 360-degree buyer account view with:
- Monetary information/insights in regards to the account
- Aggressive information panorama
- Product consumption by product line and cloud
- Buyer help circumstances
- High Use Circumstances Driving Income
- Vendor Suggestions on different use circumstances which are supplied to comparable clients
Knowledge hygiene alerts
- Use circumstances which are going stay within the subsequent week/month/quarter
- High use case blockers
- Use circumstances that lack key data (ie exec enterprise sponsor and so on.)
Gross sales collateral
- Gross sales playbooks
- Aggressive collateral
- Assembly summarization
- Pitch decks
Orchestrate motion
- Replace CRM with the subsequent steps on particular alternatives or use circumstances
- Draft a prospecting electronic mail to a brand new buyer contact
- Create a customer-facing proposal
Our discipline AI assistant answer is constructed completely on our Databricks tech stack. It permits integration into a number of and various information sources and supplies a scalable infrastructure framework for information retrieval, prompting, and LLM administration. It’s constructed utilizing the compound AI agentic framework and helps the addition of a number of instruments (SQL queries, Python capabilities) which are all ruled via our Unity Catalog governance layer.
Agent / Instrument Framework
Human inputs are inherently ambiguous; LLMs have now given us the flexibility to make use of context to interpret the intent of a request and convert this into one thing extra deterministic. To service the request, it could be essential to retrieve particular info, execute code, and apply a reasoning framework based mostly on beforehand discovered transformation. All of this data should be reassembled right into a coherent output that’s formatted appropriately for whomever (or no matter) will devour it.
That’s precisely what the sector AI assistant does to reply to the queries from the sellers. The sphere AI assistant has 1 driver agent and a number of instruments and capabilities that carry out the deterministic processing.
- Knowledge basis: That is the set of information sources that the agent interacts with. In our answer, this information basis consists of information in our Lakehouse, gross sales collateral, Google docs in addition to information that resides in our CRM (Salesforce).
- Deterministic processing: The set of capabilities and instruments required to supply appropriate, high-quality responses. The LLM can extract fields from a question and move these to an ordinary perform name to do deterministic processing. Inside the Databricks Platform, the Mosaic AI Instruments and Features capabilities allow this and user-defined capabilities can carry out most actions inside Databricks. These could be usually Python capabilities or easy SQL queries or APIs that combine with exterior apps comparable to Glean, Perplexity, Aha and so on. and these could be invoked utilizing pure language.
- LLM fashions: We leverage Azure OpenAI, GPT 4 because the foundational mannequin for the sector AI assistant answer. That stated, the framework helps a multi-model strategy the place the precise capabilities of every mannequin is evaluated with respect to the way it offers with particular use circumstances. For e.g. now we have evaluated our answer with numerous open supply fashions and we selected Azure Open AI – GPT 4 because the mannequin for our answer based mostly on the groundedness of the mannequin, its capacity to generate factual and related content material, its capacity to choose the precise user-defined perform / instrument for processing every immediate, and its capacity to stick to the content material output formatting prompting offered to the mannequin.
That stated, our answer structure is designed to permit for flexibility in adopting new fashions as they develop into out there in our Mosaic AI agent framework.
At Databricks, now we have leveraged the Mosaic AI Agent Framework which makes it simple to construct a genAI utility like the sector AI assistant. Utilizing this framework, now we have outlined analysis standards and we leverage LLM-as-a-judge functionality to attain the appliance responses. The Mosaic AI Gateway supplies entry controls, price limiting, payload logging, and guardrails (filtering for system inputs and outputs). The gateway provides the person fixed monitoring of operating methods to observe for security, bias, and high quality.
The parts that we leveraged for our discipline AI assistant are:
Answer Structure
Our Learnings
Knowledge is messy – Leveraged Lakehouse, iterative enlargement of datasets, centered on data-engineered pipelines and constructing clear, GOLD Single Supply of Reality datasets
Measuring ROI is tough – Be ready to experiment with small focus teams within the pilot. Constructing analysis datasets for measuring mannequin effectiveness is tough and requires centered effort and a method that helps fast experimentation
Knowledge and AI Governance is a MUST – Have interaction early with Enterprise Safety, Privateness, and Authorized groups. Construct a robust governance mannequin on Unity Catalog for the information in addition to the brokers and instruments
Conclusion
By this submit, we hope you discovered about our Databricks on Databrick’s GenAI journey and the way we leverage know-how like this to assist our sellers be simpler. Using GenAI for this use case has helped to showcase how AI brokers can considerably remodel and help each side of the vendor journey, from prospecting and buyer insights retrieval, driving higher information hygiene by automating repetitive guide duties and actioning these information insights to driving alternatives and bettering gross sales velocity.
Keep tuned for our upcoming posts, the place we’ll proceed to share our experiences on how AI is reshaping the vendor expertise at Databricks.
