Future, AI & Salesforce with Mohith Shrivastava
In this episode of TAGS Podcast, I had the pleasure of speaking with Mohith Shrivastava, a Developer Advocate at Salesforce, about his journey—from winning a Salesforce hackathon in 2015 to leading developer enablement at Salesforce today.
We discussed AI, the future of Salesforce development, how LLMs (Large Language Models) are reshaping coding, and why curiosity is the most important skill for developers today. If you’re a Salesforce developer, consultant, or just tech-curious, this is a must-read!
The Hackathon That Started It All
Rewind to 2015, when Salesforce rapidly evolved and hackathons were all the rage. Mohith and his team won a Salesforce Summer of Hacks event with a Sales Booster prototype—an early AI-driven automation tool before AI was mainstream!
đź’ˇ The core idea? Automate Salesforce workflows like lead management, sales operations, and knowledge sharing to reduce human intervention and enhance productivity.
Little did he know, the concept of meta-automation—which he casually coined back then—would become mainstream a decade later with AI agents and automated workflows.
From Consulting to Developer Advocacy
Mohith’s journey took him through Big 4 consulting giants, where he mastered the enterprise tech stack. But something felt repetitive:
“I was solving the same business problems with different objects, writing the same triggers, setting up the same flows.”
🚀 The Pivot: He started working with Salesforce ISVs (Independent Software Vendors), which introduced him to product thinking—how to build reusable solutions instead of custom one-off projects.
But the real turning point?
➡️ The realization that enablement and education could impact more developers than just building software for a single company.
This led him to Salesforce Developer Advocacy, where he now enables thousands of developers worldwide.
AI, Agents, and the Future of Salesforce Development
The conversation naturally shifted to AI, Agentforce, and the rise of AI-powered development tools.
“No, but developers who use AI will replace those who don’t.”
Mohith strongly believes that understanding AI’s boundaries and learning how to debug and refine AI-generated code will be the key differentiator.
âś… What Developers Should Focus On Now:
Mastering AI tools (ChatGPT, Claude, Agentforce, Gemini)
Understanding System Architecture & Prompt Engineering
Becoming code reviewers rather than just coders
Learning RAG (Retrieval Augmented Generation) for smarter AI applications
📌 Big Takeaway: Don’t just copy-paste AI-generated code. Understand it. Debug it. Improve it.
The Role of AI in Customer Experience
We also discussed how businesses are approaching AI. Many companies are sceptical about AI agents, but Mohith predicts that every business will eventually have an AI agent—just like websites and mobile apps today.
🎯 Example: Imagine a bank that instantly tells you which credit card has lounge access instead of making you dig through 10 pages of terms & conditions.
đź’ˇ Customer Takeaways:
✅ AI isn’t replacing humans—it’s replacing bad experiences.
âś… Businesses investing in AI agents will win the customer experience battle.
Choosing the Right LLM for the Job
With so many AI models, which one should developers use? Mohith shared a quick guide:
| Use Case | Best AI Model |
|---|---|
| Coding & Salesforce Development | Claude Sonnet 3.5 / GPT-4.0 |
| Long-form writing & blogs | ChatGPT (GPT-4.0) |
| Multimodal input (text, images, screen sharing) | Gemini 2.0 |
| Deep reasoning & complex problem-solving | GPT-4.o1 |
Becoming a Salesforce Developer Advocate
One of the most exciting parts of the conversation was how to become a Developer Advocate.
âś… Key Skills Needed:
Curiosity & Lifelong Learning – You must enjoy digging deep into new tech.
Technical Storytelling – You should be able to explain complex topics simply.
Soft Skills – Content writing, video editing, live streaming, presenting.
Networking & People Skills – You’ll work with developers, product teams, and customers.
💡 Myth Busted: Developer Advocacy isn’t just about “presenting slides.” It’s about building content from scratch, experimenting, and answering real-world developer problems.
Mohith’s parting advice?
đź“Ś Developers need to be more curious and creative!
📌 “Too many people are just reproducing Trailhead content instead of exploring unique use cases. AI can’t replace creativity—your unique ideas and approach matter!”
🚀 So, challenge yourself! Build something unique!
Other Podcast Episodes
References
Latest from Our Research
Queries? Lets talk
