SaaS Sessions

S9E8 - Unified Commerce & AI ft. Jigar Dafda, CPTO at Fynd

In this episode of the SaaS Sessions podcast, Sunil Neurgaonkar sits down with Jigar Dafda, Chief Technology & Product Officer at Fynd, to unpack how AI is fundamentally reshaping e-commerce in India.From conversational commerce and hyper-personalization to autonomous back offices and AI-driven customer support, this conversation cuts through the hype to explain what’s actually changing, what’s overblown, and what founders must build for if they want to survive the next decade of commerce.Key Takeaways -1. Commerce Is Shifting From Interfaces to Conversations-Traditional storefronts and search-driven UX are being replaced by conversational buying surfaces.- SEO is giving way to GEO (Generative Engine Optimization) as ChatGPT-like interfaces become the new entry point.- Merchants will still own fulfillment and data—but discovery will increasingly happen outside their websites.2. Hyper-Personalization Is No Longer Optional—It’s Infrastructure- Customer Data Platforms (CDPs) are the backbone for AI-driven personalization across online and offline channels.- AI enables real-time personalization without armies of data scientists or analysts.- The real win isn’t better targeting—it’s higher conversion with less customer effort.3. Dynamic Pricing and Forecasting Are Moving Into the Back Office- Pricing, inventory planning, and demand forecasting are becoming autonomous systems.- Decisions that once took days (via SQL and dashboards) now happen in real time.- AI shifts teams from “executors” to “validators” of system-generated decisions.4. Customer Support Is the Lowest-Hanging AI Opportunity- 60–80% of customer queries are repetitive and easily automated.- AI agents now deliver 24/7, multilingual, context-aware support at scale.- The real challenge is no longer conversation—it’s clean integration across OMS, WMS, and logistics systems.Lightning Round Insights:- Fastest way to learn today: Use ChatGPT as a personalized tutor—summarize, question, and iterate.- Hardest leadership lesson: Systems are easy. People are not.- Founder advice: Build for where the market is going, not where it is today—today’s solution will expire faster than you expect.About Fynd:Fynd is one of India’s leading unified commerce platforms, powering brands across online, offline, marketplaces, and quick commerce. From storefronts and PIM to OMS, WMS, and omnichannel integrations, Fynd enables end-to-end retail operations on a single stack.Chapters:00:10 – Introduction00:50 – Jigar’s decade-long journey at Fynd05:20 – AI before vs after ChatGPT08:10 – Conversational commerce & GEO13:40 – Hyper-personalization and CDPs19:40 – Dynamic pricing and demand forecasting30:30 – AI in customer support37:20 – Predictions for the future of e-commerce39:40 – Lightning roundVisit our website - https://saassessions.com/Connect with me on LinkedIn - https://www.linkedin.com/in/sunilneurgaonkar/

S9E7 - From Clicks to Conversations ft. Tom Drummond, Founder & Managing Director at Heavybit Industries

Tom Drummond, founder and managing director at Heavybit Industries, dissects the seismic shift from graphical interfaces to conversational AI—and why most SaaS companies are thinking about it wrong. This isn't about adding chatbots to legacy products. It's about fundamentally rethinking what software means when voice becomes the primary interaction mode and agents become your customers. Drummond reveals why rapid execution trumps defensibility, how API-first companies will dominate the agent economy, and why trust—not technology—remains the ultimate bottleneck.Key Takeaways:1. Large Language Models Are Creating the Biggest Interface Shift Since the GUIThe fundamental breakthrough isn't AI intelligence—it's natural language interpretation that frees computing from screensThis democratizes access beyond anything the mobile revolution achieved, making software ubiquitous without requiring visual interfacesThe transition will be generational: today's builders remain skeptical while the next generation will trust AI as naturally as we trust Wikipedia2. Visual Interfaces Won't Disappear—They'll Become Trust MechanismsGUIs will shift from primary interaction mode to verification and confidence layer for voice-driven commandsHumans need visual confirmation not because they're old-fashioned, but because they require ownership and accountabilityThe winning pattern: conversational interfaces for input, visual interfaces for trust and evidence3. Software Must Do More When Friction DisappearsIf you previously delivered value by creating forms to fill out, you've been selling toil—not solutionsWhen a 22-second phone call replaces a 10-minute form, your platform value collapses unless you deliver actual outcomesThe new bar: integrate across systems, automate end-to-end workflows, and eliminate entire categories of work4. Momentum Is the Only Moat That Matters at Early StageDefensibility discussions are premature when the entire market is being redefined and expectations resetInvestors have shifted to momentum-based evaluation even at seed stage—growth velocity trumps articulated competitive advantagesExecute aggressively with AI internally before worrying about competitive positioning: ship faster, operate leaner, reimagine everythingLightning Round Insights:Learning accelerator: Experiential over theoretical—don't read about programming, program; don't study sales, sellCareer wisdom: Agency and aggressive opportunity capture matter more than perfect strategy in high-change environmentsDaily AI tool: Continue (Heavybit portfolio company)—enables rapid internal tool development with model-agnostic approach and custom data sourcesChapters:00:00 - Introduction and Heavybit's Investment Thesis04:33 - The Language Model Revolution: From Screens to Voice12:13 - Rethinking Product Design: Phone Numbers Over Websites21:36 - Key Challenges: Trust, Pricing, and Agent Distribution30:04 - Differentiation Strategy: Momentum Over Moats36:56 - Why Switching Costs Still Matter for Sustainable Growth40:04 - Lightning Round and Closing ThoughtsVisit our website - https://saassessions.com/Connect with me on LinkedIn - https://www.linkedin.com/in/sunilneurgaonkar/

S9E6 - The Future of Intelligent Agents in SaaS ft. Bhaskar Roy, Chief of AI Products and Solutions at Workato

Bhaskar Roy, Chief of AI Products and Solutions at Workato, breaks down the fundamental shift from app-centric thinking to agent-first architecture. This conversation cuts through the AI hype to reveal how enterprises are moving beyond simple content generation to deploy agents at the core of business processes—from quote-to-cash workflows to IT helpdesk operations. Roy reveals the critical gap between low-agency experimentation and high-value automation, outlines the governance frameworks required for autonomous agents, and explains why getting hands-on is the only way to separate signal from noise in the agentic era.Key Takeaways1. The Agency Gap: Most Companies Are Stuck in Low-Value TerritoryGartner research shows most companies remain in low-agency activities like content creation and email writing, while maximum ROI comes from agents executing core processes like order-to-cash and supply chain managementSaaS is deterministic and requires months for customization, while agents adapt and execute in real-timeThe transformation eliminates bottlenecks where process experts depend on developers for integrations2. Purpose-Specific Agents with Enterprise Skills Beat General-Purpose ApproachesAgents must be narrowly focused with defined skills—finance agents restricted to finance users, each agent limited to relevant capabilities onlyGovernance controls agent-to-agent collaboration, with supervisor agents coordinating and ensuring complianceAgent authentication ensures agents respect user permissions—only updating data the individual user can access3. App Events and Enterprise Acumen Enable Proactive IntelligenceAgents listen to events across systems—new hires, leads, tickets—then reason, contextualize, and take autonomous actionEnterprise acumen monitors KPIs continuously and proactively recommends actions as metrics fluctuateThe shift from reactive to proactive requires human-in-the-loop now, but full autonomy approaches rapidly4. The Mindset Shift: From App-Centric to Agent-First ThinkingStop thinking about which app to use—start asking which agent can handle core business processes betterWhen agents become core to business, reliability matters—platforms must provide innovation with enterprise-grade trustReading about AI isn't learning—build agents, test them, discover what's real versus hypeChapters:00:00 - Introduction00:46 - Journey at Workato: From Integration to Intelligent Agents03:40 - The Evolution from SaaS to Agentic Workflows07:51 - Best-of-Breed Apps Creating Data Silos and Integration Challenges10:54 - Employee Productivity Gains Through Unified Agent Interfaces13:20 - The Agency Gap: Low-Value vs. Core Business Process Automation17:32 - Real-World Agent Deployments: Quote-to-Cash and IT Helpdesk20:12 - Proactive vs. Reactive Agents: Where Does Automation Stop?24:42 - Trust, Governance, and Control in Autonomous Agent Systems28:42 - Preparing for the Apps-to-Agents Transition33:04 - Lightning Round: Getting Hands-On to Shorten Learning CurvesVisit our website - https://saassessions.com/Connect with me on LinkedIn - https://www.linkedin.com/in/sunilneurgaonkar/

S9E5 - Building AI Into SaaS the Right Way: Lessons from Megh Gautam, Former CPO at Crunchbase

Every SaaS company is racing to “add AI,” but most are doing it wrong. In this episode, Megh Gautam, Former Chief Product Officer at Crunchbase, reveals the hard truths behind building AI into established SaaS products. From avoiding hype-driven features to building trust through data quality and transparency, Megh shares how Crunchbase rolled out AI-powered capabilities without breaking user trust. He also breaks down the internal alignment, cross-functional execution, and relentless feedback loops required to ship AI features that actually matter.Key Takeaways -Start with Real User ProblemsAI should not be an “add-on story” — it must solve a core customer pain.Crunchbase began with AI in search, a high-usage, high-friction feature.Prioritize critical workflows over “nice-to-have” gimmicks.Data Quality Determines TrustBad data in = garbage out, especially with AI models.Crunchbase spent a decade building clean, reliable data pipelines before layering AI.Trustworthy results require grounding AI outputs in verified “truth sets.”User Trust Demands TransparencyCustomers don’t just want answers — they want to know how those answers were derived.Explainability and confidence thresholds are essential for adoption.If unsure, don’t hallucinate — caveat results and suggest alternatives.AI is a Company-Wide Effort, Not Just a Product LaunchDesigners, engineers, PMs, marketing, and GTM must move in lockstep.Pricing, packaging, and positioning are as critical as the technical build.Internal discomfort is normal — priorities will shift faster than in traditional SaaS launches.Continuous Feedback Loops Drive IterationEarly adopter programs and dense customer feedback cycles are critical.Patterns of confusion often surface only after repeated customer interactions.AI workflows blur traditional SaaS team boundaries — ownership must evolve.Chapters: 00:10 - Introduction 00:50 - Megh’s SaaS journey (Twilio, Dropbox, Crunchbase) 02:45 - AI hype vs. solving real user problems 06:05 - Why Crunchbase started with AI in search 10:17 - Data quality as the foundation for trustworthy AI 15:07 - Overcoming AI skepticism with transparency 20:01 - Aligning product, engineering, marketing, and GTM on AI launches 25:46 - Feedback loops and customer education 30:32 - Lightning Round: Megh’s favorite AI tools 36:27 - Closing thoughts and key remindersVisit our website - https://saassessions.com/Connect with me on LinkedIn - https://www.linkedin.com/in/sunilneurgaonkar/

S9E4 - From Community to Capital ft. Max Altschuler, General Partner at GTMfund

In this powerhouse episode, Max Altschuler - founder of Sales Hacker, former VP of Marketing at Outreach, and now General Partner at GTMfund - breaks down what it takes to build and scale B2B SaaS companies with speed and substance. Max shares his journey from being an early-stage operator to becoming a SaaS investor, unpacking the real mechanics behind GTM strategy, community-driven growth, category creation, and what founders often get wrong. This is a masterclass in strategic execution and long-game thinking for SaaS leaders.Key Takeaways –1. Betting on Founders: What Drives Early-Stage InvestingFounder-market fit trumps product perfection - invest in people who are machines.Space matters: Great founders + a growing wave = potential breakout.Outreach wasn’t a spreadsheet decision - it was a conviction bet on Manny Medina's obsession and clarity.2. Strategic Distribution Is the New ProductDistribution is where the new edge is; AI levels the playing field for building.GTM is the real differentiator - your ability to generate demand at scale will define outcomes.Sales Hacker’s distribution muscle supercharged Outreach’s category dominance.3. Acquisitions Are Not Just About Revenue, They’re About VelocitySales Hacker wasn’t bought for its revenue but to compound Outreach’s strategic value.Community-led growth isn't a tactic; it’s an ecosystem asset.Integration didn’t mean conversion - it meant amplification without breaking trust.4. Lessons in SaaS Scale and GTM ExecutionDon’t underinvest in product expansion; build or buy fast when signals are clear.Self-serve isn’t optional anymore; it’s mandatory for market-wide coverage.Be obsessed, but don’t burn out your team; clarity in communication scales leadership.About the GTMfund: GTMfund is an early-stage venture fund backed by 300+ of the most respected go-to-market leaders from companies like Salesforce, Zoom, and Snowflake. It focuses exclusively on backing SaaS startups with strong GTM execution and strategic potential.Connect with Max Altschuler | Check out GTMfundVisit our website - https://saassessions.com/Connect with me on LinkedIn - https://www.linkedin.com/in/sunilneurgaonkar/