Build or Buy AI: Make the Right Call for Your Enterprise
Should your enterprise build or buy AI capabilities? This session cuts through the hype to explore the strategic, financial, and technical factors driving the decision. Expert panelists share real frameworks used by Fortune 500 companies facing this critical choice.
With Vernon Keenan & Preetam Joshi
- 1
Default to buy for most generative AI use cases on Salesforce. Only build your own model when AI is critical to core business IP and competitive advantage, such as fraud detection in insurance.
- 2
The LLM architecture is deceptively simple in concept but extremely complex in production. Building it yourself requires data engineering, ML expertise, GPU infrastructure, labeling pipelines, and ongoing maintenance.
- 3
Prompt engineering is the most accessible and impactful skill for Salesforce teams adopting AI. Zero-shot and one-shot approaches require no model training and can be packaged into buttons for end users.
- 4
Security has no industry-standard checklist yet for LLM deployments. Evaluate six areas: data protection, regulatory compliance, model transparency, access controls, vendor security posture, and AI-specific risks like bias.
- 5
Guard against prompt injection, hallucination, data leakage, and model drift as ongoing operational concerns. Use data masking, lower temperature settings, and never over-rely on model output without human review.
Build only when the AI capability is core to your business IP and gives you a competitive advantage, you already have clean organized data, and you have in-house data science talent. For everything else—email generation, case summaries, sentiment analysis—buy a solution.
Prompt engineering shapes model behavior through instructions and examples in the prompt without changing the model's weights. It is fast and sufficient for most Salesforce use cases. Fine-tuning retrains the model on your labeled data, changing its weights for domain-specific precision. It requires ML expertise and significant ongoing maintenance.
Key risks include prompt injection, data leakage of PII to third-party models, hallucinated outputs presented as fact, and sensitive data appearing in model responses. Mitigate by masking data before it leaves Salesforce, using a gateway architecture with access controls, and keeping a human in the loop.
Your Salesforce org calls an external AI model via API through a gateway layer. The gateway handles prompt engineering, data extraction and masking, and sends anonymized data to your chosen cloud provider. The response is de-masked and presented back to the user, with all access governed by Salesforce permission sets.
Model drift occurs when the characteristics of your data change over time, causing a previously well-performing AI model to produce degraded outputs. Implement observability by periodically running known inputs against expected outputs, and version-control your prompts like production code.
Why AI Pilots Fail—and How to Pick the Use Cases That Don't
Most AI pilots fail because teams chase the wrong use cases. This session breaks down a simple framework for identifying high-value, immediately achievable AI opportunities that don't require months of training data—and how to avoid the hype trap that killed blockchain, web3, and every other trend.
Ian Gotts & Saurabh Gupta · 12 min watch
Build Your Salesforce AI Roadmap: The Crawl-Walk-Run Framework
The final session of Enterprise Dreaming synthesizes everything you've learned about AI into an actionable roadmap. Using the crawl-walk-run framework, learn how leading enterprises move from experimentation to scaling AI across sales, service, and customer-facing operations.
Anand B Narasimhan & Saurabh Gupta · 12 min watch
AI-Driven Personalization Without Breaking Compliance in Financial Services
Financial services leaders face a paradox: deliver Tesla-Starbucks-Apple-level personalization while managing GDPR, CCPA, and regulatory scrutiny. This session walks through concrete use cases—advisor co-pilots, sentiment analysis on claims, account 360 views—showing how to mask sensitive data before feeding it to AI, maintain compliance, and actually keep customers.
Kavin Mehta & Saurabh Gupta · 11 min watch
One email a week. Practitioner analysis, no vendor spin.
The Weekly Brief — sessions and essays from senior Salesforce practitioners working through AI, security, and architecture in production.
One email each Friday. The sharpest analysis on Salesforce, AI, and enterprise architecture — written for senior practitioners.