ICCD Holding
dotpy.tech Delivers Two-Day AI Program to ICCD Holding
A hands-on executive program taking ICCD Holding's teams from AI fundamentals for decision-makers to building their own working AI agents — in two days.
In September 2026, dotpy.tech delivered a two-day corporate AI program for ICCD Holding, bringing together managers and specialists from across the organization's functions inside the company's own boardroom. The agenda was deliberately split in two: one day to build the managerial judgment needed to lead AI adoption, and one day to put working AI agents in participants' hands. By the close of day two, every attendee had moved from discussing artificial intelligence to operating it.
Two days of hands-on AI training with the ICCD Holding team — September 2026.
Across the region, the conversation about artificial intelligence has shifted decisively. Organizations are no longer asking whether AI belongs in their operations; they are asking who inside the building is equipped to deploy it, govern it, and defend the decisions it informs. ICCD Holding commissioned this program to answer that question directly — not with a keynote, but with two days of structured, practical work.
dotpy.tech designed the curriculum around a single principle that governs all of its corporate engagements: no participant should leave a training room with knowledge they cannot use on Monday morning. Every concept introduced over the two days was paired with a live demonstration, a business scenario drawn from the participants' own functions, and an exercise they completed themselves.
AI for Managers
The first day addressed the audience that determines whether an AI initiative succeeds or stalls: the managers who approve it, resource it, and answer for its results. Rather than opening with technology, the session opened with decisions — where AI genuinely creates leverage inside an organization, where it quietly creates risk, and how a manager distinguishes a serious use case from an expensive distraction.
Participants worked through the practical anatomy of modern AI systems: what large language models actually do, why they produce confident errors, and what that means for any workflow where accuracy is non-negotiable. dotpy.tech's trainer demonstrated live how the same business task — drafting a report, summarizing a dense document, extracting structure from unformatted data, interrogating a spreadsheet — is transformed when delegated correctly, and how it degrades when delegated carelessly. The emphasis throughout was on prompting as a managerial skill: specifying context, constraints, and output format with the same precision a manager would use when briefing a capable new hire.
The day closed on adoption and governance. Teams mapped candidate use cases from their own departments against effort, risk, and expected return, then examined the guardrails that make AI usable in a professional setting — what should never be pasted into a public model, how outputs must be reviewed before they reach a stakeholder, and how to write a simple internal policy that enables staff rather than paralyzing them. Participants left with a shared vocabulary for evaluating AI proposals and a realistic sense of what the technology can and cannot be trusted to do.
Building Mini Agents with Claude
Day two moved participants from directing AI to building it. The focus was the agent: an AI system given a defined role, a set of instructions, and the reference material it needs to perform a recurring task reliably — the difference between asking a model a question and deploying an assistant that handles the same request the same way every time, for every member of a team.
Working in Claude, participants built mini agents from the ground up. The session broke the craft into its components — writing a role definition that constrains behavior rather than describing a personality, supplying the knowledge an agent must draw on, specifying output structure so results arrive ready to use, and handling the edge cases where an agent should decline or escalate instead of improvising. Each participant built and tested an agent tied to their actual responsibilities, then refined the instructions as the failure points surfaced, learning the iteration loop that separates a demo from something a department can rely on.
The day also covered the practical scaffolding around these systems: where an agent belongs in an existing process, how to document and share one so colleagues use it consistently, and how to recognize when a task warrants a purpose-built agent versus an ad-hoc prompt. By the closing session, the ICCD Holding team had working assistants of their own creation — and, more valuably, a repeatable method for producing the next one without external support.
“We came in expecting a lecture about AI. We left with tools we had built ourselves.”Training Participant · ICCD Holding
Key Outcomes
What the ICCD Holding team took away from the two days:
- A shared framework for identifying high-value AI use cases and rejecting low-value ones
- Practical command of prompt engineering applied to real managerial and departmental tasks
- Each participant built and tested a working mini agent in Claude tied to their own role
- A repeatable method for designing, documenting, and refining new agents independently
- Clear guardrails for accuracy, confidentiality, and review before AI output reaches stakeholders
- A prioritized shortlist of departmental automation opportunities to pursue next
Partner with us for your next corporate AI training
dotpy.tech is an AI and Data Science training academy working with organizations across Egypt and the wider MENA region. We design and deliver corporate AI programs, executive briefings, prompt engineering and automation workshops, and full AI diplomas — built around the client's own workflows, delivered in English or Arabic, and measured by what teams can do afterward rather than what they were shown.
Every engagement is scoped to the client: we audit where AI creates real leverage in the business, build the curriculum around those use cases, and train the people who will own them.
Request a training proposal →© 2026 dotpy.tech — AI & Data Science Training Academy. Program delivered for ICCD Holding, September 2026.