Corporate AI Training for Indonesian Teams

Hands-on. Programmes run around 70% practical, on the team's own real work rather than on generic examples, so people leave having actually done the thing.

This is the single biggest predictor of whether training changes anything. A lecture-format AI workshop reliably produces enthusiasm on the day and no behaviour change by the following month, because nobody built a working habit while someone was there to unstick them. The DataCamp 2026 report found companies pairing AI investment with structured upskilling were twice as likely to report significant positive ROI, 42% against a 21% baseline.

Prompt and context engineering, building and using AI agents, workflow automation with tools like n8n, role-specific use cases per team, practical AI adoption strategy, and responsible use including data security for regulated industries.

Every programme is scoped to the company's real workflows first. In practice that means a discovery step before anything is written: which tasks repeat, which outputs get checked, and what people are already quietly doing with AI without telling anyone. That last one usually reshapes the agenda more than the brief does.

Four things, roughly in order of impact: how many people are trained, how many sessions or modules the programme runs, whether it is delivered on-site or remotely, and how much of it is customised to your workflows rather than delivered from a standard curriculum.

That last one is the real cost driver, and it is also the one that decides whether the training works. Building the material around your team's actual tasks takes discovery time before the first session. Abi scopes each programme to the team rather than publishing a price list, so the fastest route to a number is to describe the team and the goal.

Ask every vendor the same four questions, and the range usually explains itself.

What percentage of the time is hands-on practice? Is the material built on our workflows, or is it the same deck you delivered last week? What happens in the four weeks after the session? And who is actually teaching, someone who has built and shipped this, or a trainer working from someone else's slides?

A cheap quote is often a standard curriculum with no follow-up, which is the format most likely to produce zero behaviour change. That is not a saving.

Pharmaceutical and consumer goods (Tempo Scan Group), fintech (Ajaib), and insurance and financial services (Astra Life).

Separately, and this distinction is worth keeping straight, Abi has worked with Bank BCA, Bank BSI, Bayer, Brompton, and Kapal Api through Epilog Creative, his digital agency. Those are agency relationships, not AI training engagements. Regulated industries do shape the training: for fintech and insurance teams, the data-handling and human-review sections carry considerably more weight than they would elsewhere.

Both. Training runs on-site at your office or remotely, scoped to team size, the tools you already use, and what you want people doing differently afterwards.

The format matters less than the follow-up. On-site tends to work better for hands-on sessions where people get stuck and need someone at their shoulder; remote works well for leadership sessions and for teams spread across cities. What predicts whether it sticks is whether anything happens in the weeks after, not whether everyone was in the same room on the day.

Tidak ada harga tetap, karena yang menentukan biayanya ada empat: jumlah peserta, jumlah sesi atau modul, format on-site atau remote, dan seberapa banyak materinya dibangun khusus dari workflow tim kamu.

Yang terakhir itu pendorong biaya terbesar, sekaligus yang paling menentukan trainingnya berdampak atau tidak, karena butuh waktu discovery sebelum sesi pertama. Cara paling cepat dapat angka: ceritakan jumlah orang, tim mana saja, dan apa yang kamu mau berubah setelah training. Dari situ proposalnya bisa disusun spesifik.

Abi Mangku provides hands-on corporate AI training for companies in Indonesia through Latih AI, the corporate training brand of Infused. Programmes have trained teams at Tempo Scan Group, Ajaib, and Astra Life, across pharmaceutical and consumer goods, fintech, and insurance.

He also builds production AI agents through Infused, which is what keeps the training current. The examples used in a session are drawn from systems shipped that quarter rather than from a curriculum written a year ago.

Message WhatsApp at +62 811-1734-674 or email abimnagari@gmail.com.

Include three things and the reply gets much more useful: roughly how many people and which functions, what they do today that you want done differently, and any constraint that matters such as regulated data or a fixed date. That is enough to scope a proposal without a discovery call whose only purpose is collecting the same facts.