"3 months ago I didn't know the difference between an AI agent and a chatbot."
Suzanne Jana Chew wrote that in a LinkedIn post looking back on her internship with us at Anchor Sprint. By the end of it she had built an extended content calendar and presented a marketing strategy to our team.
Suzanne Jana Chew
Product Growth Intern, Anchor Sprint
University of Nottingham Malaysia
Suzanne, a student at the University of Nottingham Malaysia, joined us as our Product Growth Intern: an AI internship in Malaysia at an agentic AI development agency in Shah Alam, Selangor.
She published the full story on 21 September 2026. Read Suzanne's own account on LinkedIn.
A content calendar built around the reader
Her core work was planning and drafting an extended content calendar spanning carousels, reels, posters, infographics and blog posts.
It was built to walk a reader who has only vaguely heard of AI to the point of being ready to book a consultation. That meant real decisions about which topic belonged at which stage.
Owning a strategy end to end
Suzanne also researched and presented an SEO and SEM (search engine optimisation and paid search marketing) session to the wider team. It was her first time owning a strategy end to end.
She described the questions afterwards as sharp, but never designed to catch her out.
What she ended up explaining
Suzanne learned the product by explaining it:
- why an agent is more than a chatbot
- how it plugs into a company's tools through MCP (the Model Context Protocol)
- why an agent gets only the same access an employee already has, with no master key
Her own summary of how that went:
"Writing about agentic AI for other people turned out to be the fastest way I've ever learned a technical topic myself."
Her article names what she sees as the interesting question: "exactly where the line should sit between what an agent can safely decide and what still needs a person."
If you want the plain-language version of the MCP part, start with how AI connects to your business systems.
Learning by doing the work is the same point we make in why AI training often does not reach the job: a course alone rarely changes how the work gets done.
What our CEO said in reply
Jazz Tong, our Co-Founder and CEO, replied on LinkedIn to thank her. He wrote that the team was still using both the content calendar and the SEO and SEM strategy she presented.
He also wrote:
"You came in with no technical background, and by the end you were explaining MCP, agent permissions and where the human-in-the-loop line should sit."
One person's account
Suzanne closed her article on three things: real work, real trust, and people invested in helping her understand.
We cannot promise every internship will look like hers. This is one person's account of three months here.
If you are looking for an AI internship in Malaysia, see what is open on our careers page. Roles there are listed as hybrid, based in Shah Alam, Selangor.
If no listed role fits and you want a marketing or product growth internship at an AI company in Malaysia, you can introduce yourself anyway.
Free consultation
Want to learn agentic AI by doing the work?
See which roles and internships are open at Anchor Sprint, or introduce yourself if none fits yet.
Sources
- Suzanne Jana Chew, What Three Months at Anchor Sprint Actually Taught Me (LinkedIn article, 21 September 2026)
- Suzanne Jana Chew, LinkedIn post introducing the article
Quoted with Suzanne's permission.