Summit Strategy: Key Considerations When Implementing an AI-First Approach
Adopt an AI-first mindset: Align the overall implementation timeline and planning with an AI-first mindset.
Basic Setup: Once the primary Sales and Industry Cloud setup is in place, evaluate sales-specific AI use cases in the MVP phase.
Initial Use Cases: Start with predictive AI use cases such as lead scoring, Einstein Activity Capture, opportunity scoring, and Einstein forecasting.
Seamless Integration: Integrate AI initiatives seamlessly into the implementation cycle rather than running them as separate projects.
Opportunity Identification: Coordinate with business and process SMEs to spot opportunities for creating use cases.
Service AI Features: Gradually roll out AI-specific service features.
Initial Rollout: Implement features like Einstein case classification, article recommendation, case routing, and next best action.
Subsequent Rollout: Follow with engagement scoring, behavior scoring, content selection, and marketing insights.
Generative AI Use Cases: Transition to generative AI use cases as the organization becomes more AI-ready.
Sales Applications: Implement use cases that are generative rather than merely prediction-based, such as sales emails, call summaries, clause generation, and contract digitization.
Advanced Features: Introduce forecast guidance, work summaries, and other summarization and extraction themes.
Further Exploration: As the organization evolves into an AI-powered entity, explore generative features for code generation, test case creation, and test class generation.
Advanced AI: Investigate prompt creation, fine-tuning, and alternate models for new features.
AI Awareness and Readiness: Foster AI awareness and readiness within a small group of dedicated team members.
Dedicated Team: Support the team with access to platforms like Salesforce Einstein or similar open-source models and libraries.
Internal Introspection: Start introspection and AI implementation internally with the help of this AI-aware group.
A dedicated Center of Excellence (CoE) and AI Lab are crucial for enhancing organizational collaboration. An AI CoE can guide a typical architecture, platform, model evaluation, technology, security, trust, compliance, ethics and responsible design. Including stakeholders from different parts of the organization ensures joint decision-making regarding investments.
In an AI Lab, use cases can be created and demonstrated to assess the practicality and effectiveness of solutions rather than merely adopting AI for the sake of it. There are various approaches to accomplishing tasks with AI models, and decisions need to be made on using prompts or fine-tuning. This requires some education, which is where the CoE can be invaluable.
Infosys Salesforce practice created 'TeleBot' journey in 2019 with a different market perspective. We used sales and ordering use cases instead of the typical services/contact center use cases. Back-office requests such as number porting, plan upgrades, and capacity changes could be handled in this case. Agentforce accomplishes the same journey and delivers a better customer experience as agents can take the order from the customer in Natural Language, helping create assets in the system for, say, a 5-G data plan or new SIM purchase. All this happens while the customer speaks in Natural language to order or even place a ‘disconnect’ request for the existing assets using a handheld device. For an action added to the topic in Agentforce, the commerce API can do the rest and be applicable for SoHo, B2C, Retailers, etc. This can even be replicated in the energy segment when the customer moves home, requesting new energy connections or renewable products.
