ArticlesVeda: Certinia Veda AI Overview FAQ

This article answers frequently asked questions about Certinia Veda.

Contents 

• General Questions
• Product Functionality & Delivery 
• Certinia’s Use of AI & Your Data
• Commercials & Terms

General Questions

What is Veda?

• Veda is AI built natively inside Certinia, combining specialist agents and generative capabilities grounded in decades of institutional memory to deliver precision at every stage of the services journey

Can Veda be put into production?

• Yes, but we recommend you start by deploying Veda in a full copy sandbox environment for testing before making a production deployment
• Any org that utilizes Veda must be on the Spring ‘26 version or newer
• Orgs must have the required AI packages and Agentforce installed

Is Agentforce required to use Veda?

• Yes; access to Certinia’s Agents requires Agentforce

Can new customers use Veda immediately?

• Certinia Veda GenAI and Agents leverage data found inside of PS Cloud and CS Cloud
• For new customers deploying Veda, understand that Agent and GenAI performance will be limited at first as data is being populated in the system
  • Once the system is being used and necessary data points are populated, Certinia Veda will rapidly provide articulated value to the target users

How hard is it to adopt Veda?

• Adoption is simple
• Veda’s capabilities are embedded directly into the daily workflows your team already uses in Certinia
• It works alongside users to provide summaries and automate tasks, reducing the administrative burden rather than adding a new digital chore 

Why should I use Veda instead of building my own agent using a generic LLM or existing platform tools? 

• Connecting a home-built agent to PS Cloud via API is possible, but those API connections mostly limit the agents to simple query and summarization tasks, and do not allow for true operational agents
• Building your own agent requires solving for the "Maintenance Spiral," meaning every time a model updates or a business's logic changes, the custom-built agent is likely to break. By using Veda, you ensure your agents are always maintained with the latest version of PS Cloud
• You get the innovation without the massive technical debt of maintaining a custom AI stack 

Product Functionality & Delivery 

What AI functionality is currently available?  

• The Veda suite will expand over time
• As of the Spring ‘26 release, the available capabilities are Staffing Agent, Project Assistant Agent, and various AI-generated summaries

How is Veda different from other AI solutions?

How is Certinia AI packaged and delivered?

• To gain access to Certinia Agents or use Certinia’s Generative AI features like summaries, you must install an additional Certinia Core AI package
• For Agents specifically, you also need the relevant package that includes the Agent
  • These Agent packages contain an Agent Template used for deployment
  • Access is then provided via permission sets and appropriate agentforce licenses
  • Agents are included in separate packages to avoid adding dependencies on AI technology to the main PS, CS, and FM packages
• For Generative AI summaries, once the Core AI package is installed, the summary generation ability will appear provided users are licensed for the technology and assigned the required permission sets

How do users gain access to Certinia Actions?

• The actions utilized by the Certinia Agentforce agents are contained within the additional Certinia agent packages
• A license for the additional package is required to use those actions

Will Veda’s AI agents make staffing or project decisions without my oversight?

• No, Certinia views Veda as a "Bridge to Autonomy"
• Veda acts as a highly skilled digital assistant that performs the heavy lifting of data distillation and drafting
• While the agents can suggest staffing plans or identify project risks, human experts remain in the driver’s seat
• Veda provides the speed and the draft; the human expert provides the final approval and strategic direction

How does Veda handle the "Context Gap" that generic AI tools struggle with?

• Generic AI understands language, but it does not understand "services math"
• Because Veda is embedded directly into PSA, it has full context of project hierarchies, resource skills, and revenue recognition rules
• It does not guess; it calculates based on the hardened logic that already runs the business, ensuring that every AI-generated summary or staffing suggestion is grounded in the specific reality of the organization

Is it possible to self deploy Certinia Veda solutions?

• Self-deployment is recommended for administrators that have sufficient experience in Salesforce and Certinia products via the PS deployment guide or CS deployment guide

What is required to implement Veda? 

• Agentforce and Certinia AI license
• Certinia Spring ‘26 is installed in the testing org (a full copy sandbox to start)
• The Certinia team must install the necessary Certinia AI packages, which are not part of the regular package installations

What LLM is Certinia using in the implemented solutions? 

• For Generative AI Summaries, Certinia uses the Salesforce default LLM (currently OpenAI GPT4 Omni Mini) which is considered a standard level of LLM in the Salesforce rate card
• This may be subject to change as Salesforce deprecates and replaces models
• For Agents, Certinia uses the default Agentforce model as there is presently no ability for Certinia to choose

Is it possible to change the LLM that is being used?  

• Generative AI Summaries provide some flexibility into which LLMs can be used such as Gemini and GPT3-5
• Currently, Certinia Generative AI does not support other LLMs
• Agentforce does not provide users with a choice of LLMs

Is there a connection point to connect third party AI aside from Agentforce?  

• Certinia does not have a dedicated API connection point, but this is an area of active investigation
• Be aware of licensing implications to connecting third party AI

Does Veda require “training” or “tuning” on our specific data before it becomes accurate?

• Certinia Veda does not require traditional LLM training or manual tuning to become functional; instead, it is designed to deliver immediate value by leveraging your existing data
• However, the accuracy of Veda’s output is directly proportional to the quality of the data it consumes
• Rather than training a model, the focus is on data integrity—ensuring that your org's data is clean, structured, and accurate

What version of Certinia is required to deploy Veda?

• Customers must upgrade to Spring 2026
• Earlier versions are not supported as Generally Available product

What is the upgrade experience for Certinia Agents that have been deployed?

• Per Salesforce design, the current upgrade process involves deleting any existing agents and creating a new agent using the upgraded template
• Upon package upgrade, any predeployed Certinia agents are not modified and will not automatically learn new actions

What are the implications to the upgrade process if I customize my Certinia Agent with additional actions?

• Certinia provides Agent templates per the Salesforce Agentforce architecture
• As such, you may have to rebuild the Agents on product upgrades
  • Agent Topics, instructions, and Actions are all saved in Agentforce, but the Agent must be reassembled

Certinia’s Use of AI & Your Data

Is Certinia AI enabled on Certinia products by default? 

• No, AI powered by Agentforce or LLMs is not included in our products by default
• You will need to specifically opt-in to make use of predictive, generative, or Agentic AI capabilities
• Certinia AI is in separate, additional managed packages which are installed only when purchasing the necessary packages and licenses
• Certinia does have included features such as intelligent staffing or smart rescheduling
  • These features are intelligent in nature, but in the sense that they leverage non-deterministic logic
  • These features are included in the standard application packages though some may require additional licensing, such as CRM Analytics-based features like smart rescheduling or skills suggestions

Does the product code provide alternatives or opt in/out functionality for AI-enabled features?  

• Access to Certinia AI features/functionality is controlled primarily through user permissions and licensing
• When using Certinia’s AI package(s)along with Agentforce, you may control who can use the different AI solutions through permissions

Will product users know if they are utilizing an AI based feature? 

• Yes, Certinia Veda Agents interact with users in a separate Agentforce Pane within our UI
• For GenAI, such features are accompanied by our AI sparkle Icon
  • Use the verb "Generate" and the results come with a caution about the possible issues with GenAI

How will Certinia use my data? 

• Certinia’s policy on use of customer data aligns to the overall Salesforce policy
• The policy will be included in your Product Terms when purchasing our AI solutions

Will Certinia use my data to train the AI models?

• Certinia, or Salesforce on its behalf, does not use your data to train generative AI models 

Will Certinia send my data to any third parties for model training?  

• Your data may be made available to Salesforce for training AI models if you opt-in to such

How do we ensure our proprietary project and margin data isn't leaked into a public model?

• Veda is built on a foundation of platform trust; your data is never used to train public LLMs
• We utilize a secure, hardened architecture that ensures your institutional data remains within your governed environment
• Veda acts as a closed-loop system, applying our deterministic logic to your data to produce results that only you see

Will my data be used to create, develop, or improve AI components in the product? 

• Certinia does not get access to the data used through Salesforce AI technologies
• However, your submitted feedback and ideas will be used to improve AI components

What data privacy standards or guardrails do Certinia’s AI features abide by? How are these standards tested? 

• Zero Data Retention:
  • When your data is used in a prompt for an external Large Language Model (LLM) like one from OpenAI or Anthropic, Salesforce has a strict contractual agreement that the LLM provider cannot retain or use that data for any purpose, including training their models
  • The data is used ephemerally to generate the response and is then discarded
• Data Masking and Anonymization:
  • The Einstein Trust Layer automatically detects and masks Personally Identifiable Information (PII) and other sensitive data before the prompt ever leaves the Salesforce trust boundary to be processed by an external LLM
• Salesforce's Trusted AI Principles:
  • All AI development is guided by five ethical principles: Responsibility, Accountability, Transparency, Empowerment, and Inclusivity
  • See the Salesforce Trust Layer for more information

How do Certinia’s AI features handle permissions? 

• Certinia’s AI features respect the user's permissions
• Users cannot query or modify records they do not have access to

Commercials & Terms

Are Veda features included with my Certinia subscription?

• No, access to Certinia AI requires transacting Veda for AgentforceSKUs

Why is Veda a separate per-user/per-month charge plus consumption?

• The subscription fee provides access to current and future AI solutions - so as we add more, you get more with no incremental charge
• The consumption model ensures your costs stay aligned with the actual business value and output the AI generates for you

Do I need to be using Salesforce Agentforce to use Veda?

• To leverage Veda for Agentforce features, you must obtain Agentforce directly from Salesforce
• This includes $0 prerequisites which will come with initial usage capacity as well as additional Flex Credits based on usage

Which users need the Veda User Add-On for Agentforce licenses?  

• Licenses only need to be added for the users who are leveraging the specific Veda features including Generative AI, Agents, and Agent Actions

Is there a minimum seat requirement for Veda - must licenses be purchased for all PSA users, or can it be limited to select users such as only Project Managers?

• The User Add-On model is designed to allow you flexibility to add AI features only for those who need it

Do I need to make a commitment for consumption or usage credits?

• Obtaining Salesforce prerequisites for Agentforce will provide you with initial Flex Credits to begin using Veda features
• Generally, a large prepaid commitment is not required
• Salesforce offers a variety of options to meet your needs including pay-as-you-go, or a prepaid commitment of various sizes and types

How do I track and manage our consumption or usage credits? Do credits roll over? What happens if we run out mid-month?

• Flex Credits may be tracked in your Digital Wallet directly in the Salesforce administrative interface
• The Digital Wallet will show Flex Credit Entitlements and Flex Credit Usage
• Entitlements are tracked at the contract/sales order level, and are pooled together for overall use at the Account level
• Credits are consumed first from the earliest expiring contract
• It is not currently possible to reserve entitlements for a specific feature area

Is there a dashboard to help track consumption? 

• The usage dashboard will show Flex Credit consumption by Agent and feature
• Additionally, you can create custom tags for tailored tracking 

What happens when Certinia releases new agents? Are they bought individually?

• No, your Veda subscription grants you access to all future innovation and solutions underneath Veda
• As our R&D team rapidly develops and releases new specialist agents (for resource management, billing optimization, etc.), they are automatically made available to you
• Your subscription is essentially an "innovation annuity" that ensures your stack is always state-of-the-art

How do I measure the "hard ROI" of an AI suite like Veda?

• You will see a productivity boost for your employees around administrative workflows  
  • For example, a Project Manager spends 30+ hours a month on manual reporting and milestone tracking. Veda reduces that by up to 60%, effectively giving your PMs 20 hours of capacity back every month
• You will also have better ability to scale your project volume over time without a corresponding increase in administrative headcount  
• Greater administrative productivity will also translate into improved performance of your services business overall: faster time to staff, higher billable utilization and reduced unbilled time

Are there any other license requirements needed from Certinia or Salesforce to deploy Veda?

• The license requirements are:
  1. Existing Certinia license including required underlying permissions
  2. Salesforce Agentforce Access and Flex Credits

What licenses are needed from Salesforce?

• For the latest information - always refer to the Certinia Licensing Guide