In this episode of Voices Beyond Borders with Emma Woelk, we dive deep into the real mechanics of AI adoption inside companies — far away from buzzwords, hype, and PowerPoint visions.

My guest, Driss Farissi, founder of H-IN-Q, has spent more than 20 years in global market research and AI-driven process automation.

He explains how companies can unify feedback from all channels — WhatsApp, email, chat, phone, and social — and build a single data layer that turns scattered signals into immediate, actionable decisions.

We discuss why most digital transformation projects fail long before the tech is ready, how to overcome internal resistance, and why the smartest companies start with small pilots instead of massive roadmaps.

You’ll also learn how speech-to-text, computer vision, and automated sentiment tracking create real-time visibility for Sales, Customer Success and operational teams.

If you want to understand how to implement AI systems that actually get adopted — not just bought — this conversation gives you the pragmatic blueprint.

00:00 Intro — Why customer signals are the real foundation of AI

02:45 Driss’ background and the move from research to AI systems

08:10 How to unify WhatsApp, email, chat and social into one feedback layer

13:20 Speech-to-text, computer vision and automated analysis

18:40 Why AI projects fail: resistance, culture, and missing involvement

25:05 Starting small: low-hanging fruits and ROI-first pilots

32:10 Real-world cases: what worked and what failed

38:55 The vision: building a Single Source of Truth across the company

45:20 Final takeaways & the automation roadmap

What You’ll Learn

– How companies transform scattered customer inputs into real-time decisions

– Why a Single Source of Truth is the backbone of every AI initiative

– How speech-to-text and computer vision convert natural feedback into structured data

– How to detect early-warning signals and trend shifts automatically

– Why adoption fails when teams aren’t involved early

– How to design AI pilots that deliver ROI in days, not months

– The best starting points for automation in mid-sized organizations

– Why data culture matters more than tools or model selection

Key Discussion Topics

1 — Building a Real “Single Source of Truth”

How to consolidate WhatsApp messages, emails, chat conversations, social posts and support interactions into one data layer.

Why this matters for Sales, CS, operations and leadership — and how it enables faster, more accurate decisions.

2 — Speech-to-Text & Computer Vision in Business

How natural language and image input become structured insights using speech-to-text pipelines and visual classifiers.

Real use cases where teams shorten manual work from hours to seconds.

3 — Adoption, Resistance & Team Involvement

Why employees resist digital projects even when the tech works.

How to involve teams early, create ownership and avoid silent blockers that kill automation.

4 — AI Pilots That Deliver Fast ROI

Why starting small creates momentum.

How low-hanging-fruit automations outperform big “AI strategies” — and how to prioritize correctly.

5 — Detecting Signals Before They Become Problems

How automated feedback systems reveal trends, sentiment shifts and anomalies weeks before traditional reporting would notice.

Why this matters for churn prevention, quality control and retention.

6 — The Vision: A Unified Decision Layer

How AI-enabled companies move from manual processing to real-time insight loops.

Why this becomes the operating system for the next decade of business.

Actionable Insights

– Identify 3 customer signal sources you’re currently not analyzing

– Build a...

In this episode of Voices Beyond Borders with Emma Woelk, we explore the real, practical side of artificial intelligence in business — far beyond the hype. How can a six-person team outperform a global corporation using the right mix of automation, AI-driven sales systems, and process intelligence? My guest, Kelwin Fernandes, founder of NILG.AI and one of Europe’s most respected AI practitioners, explains how small and mid-sized companies can use AI to automate lead generation, follow-ups, PR, and content visibility inside ChatGPT and Perplexity.

We also discuss the future of AI-driven marketing, the coming energy bottleneck, and the risk of model collapse when machines start learning from synthetic data.

00:00 Intro – Emma in Porto and why sales comes first

02:40 Kelwin Fernandes on early AI work and the shift since ChatGPT

08:10 Automating lead generation and follow-ups

13:45 How AI rewrites PR and outreach strategies

18:20 ChatGPT SEO and Perplexity visibility

25:05 Agent buyers and CAC/LTV re-alignment

32:30 Model collapse explained

38:40 Energy and infrastructure challenges

44:10 Fine-tuning vs using proprietary models

50:00 Future of work – training juniors in an AI world

57:30 Final takeaways and automation playbook

What You’ll Learn

-How AI transforms sales and marketing before any other department

- Building a predictable lead pipeline with CRM automation (HubSpot, Salesforce, Pipedrive)

- Turning ChatGPT & Perplexity into the new SEO channels

- Using automation as a service without developing your own model

- Why “agent buyers” could change customer acquisition forever

- How to manage energy use and efficiency in the AI era

- Understanding model collapse and the value of human data

- The rise of micro-agencies and how individuals can scale like companies

- What every CMO and sales leader should learn right now to stay relevant

Key Discussion Topics

1 — AI in Sales & CRM Automation

Why sales is the perfect starting point for AI adoption. Learn how to automate outreach, follow-ups, and research using real-world CRM data — and how to connect those systems to LLM-based agents for measurable ROI.

2 — ChatGPT SEO and AI Search Optimization

Search is shifting away from Google. We break down how ChatGPT and Perplexity decide which websites and voices to cite — and how to position your company so your content appears inside those answers.

3 — Marketing Automation & PR at Scale

Kelwin shares how NILG.AI automates event pitching, press releases, and media outreach. The process runs 24/7 without agencies — showing what’s now possible for small teams.

4 — The Energy Problem of AI

Training AI models consumes massive energy. We unpack the real numbers, potential solutions (micro nuclear reactors, space-based energy), and how businesses can prepare.

5 — Model Collapse & Data Shortages

What happens when AIs start training on their own synthetic output? We discuss the risks and why companies like Reddit and The Washington Post are being acquired for high-quality human data.

6 — The Future of Work

As AI automates junior tasks, senior roles become strategic orchestrators of intelligent systems. We discuss how marketers and founders can evolve into decision architects, not just content producers.

Actionable Insights

- Map your CRM flow and identify 3 decisions you can automate this week

- Add AI triggers for follow-ups and newsletter engagement

- Run PR automation using AI search and event detection APIs

- Treat yourself as a micro-agency with AI agents as your team

- Track energy efficiency + data quality as new success KPIs