Buna Constructions Jeddah
dotpy.tech Delivers a 4-Day AI & Automation Transformation to Buna Constructions
Over four intensive days in Jeddah, dotpy.tech's training team partnered with Buna Constructions to take the organization's workforce on a complete journey — from the fundamentals of prompting large language models to designing and deploying autonomous AI agents capable of resolving real operational bottlenecks. The program reflects a broader wave of AI adoption sweeping across Saudi Arabia's construction and engineering sector, and Buna's leadership emerged as an early mover in putting that shift into practice.
Snapshots from the 4-day dotpy.tech × Buna Constructions training program — live sessions, hands-on automation building, and agent architecture workshops.
From Zero to Prompting: Building a Foundation in AI Communication
The program opened by grounding the Buna Constructions team in the single most important skill of the AI era: how to communicate effectively with a large language model. Rather than treating prompting as a trick or a shortcut, dotpy.tech's trainers framed it as a discipline — one with structure, repeatable frameworks, and measurable quality differences between a vague instruction and a precise one.
Participants worked through dotpy.tech's proprietary prompt engineering framework, practicing on scenarios pulled directly from Buna's own operations — procurement follow-ups, site reporting, and client correspondence — so that every exercise had immediate, tangible relevance. By the end of the day, engineers and managers who had never written a prompt beyond a simple chatbot question were producing structured, role-based prompts that consistently returned usable, professional output.
This foundation mattered because everything that followed — automation, agents, and operational problem-solving — depends on the team's ability to instruct AI systems with clarity. Day 1 ensured no participant was left behind before the program moved into more advanced territory.
AI Tools & Automation in Practice
With prompting fundamentals in place, the second day shifted toward the practical AI and automation toolkit that modern organizations rely on daily. Trainers walked the team through a curated stack of productivity and automation platforms, showing not just what each tool does, but which real Buna Constructions use case it solves best — from drafting site correspondence to summarizing lengthy technical documents in seconds.
A significant portion of the day was dedicated to Make.com, where participants watched — and then built alongside the trainers — live automation scenarios connecting forms, AI models, spreadsheets, and email in a single working pipeline. This hands-on approach demystified automation, replacing the perception that it requires a developer with the reality that a well-trained non-technical team member can build production-ready workflows themselves.
By day's end, the Buna team had a working mental map of which tool to reach for depending on the task at hand, and several participants had already sketched automation ideas specific to their own departments.
Inside the Anatomy of an AI Agent
The third day marked the program's turning point: moving from using AI tools to architecting AI systems. Trainers broke down the anatomy of an AI agent piece by piece — the LLM as the reasoning "brain," memory for context and continuity, a connected knowledge base for company-specific facts, tools for taking real actions, and system instructions that define the agent's behavior and boundaries.
Working on a shared whiteboard, the team mapped out this architecture live, translating each abstract component into something tangible for Buna's own environment — what an HR-facing agent would need to know, what a site-operations agent would need to query, and how strict, deterministic instructions differ from open-ended conversation. This session deliberately emphasized designing agents with firm behavioral rules rather than casual personalities, ensuring reliability in a business-critical setting.
By the close of Day 3, participants understood not only how AI agents work internally, but how to design one intentionally for a specific operational role — a prerequisite for the hands-on build that would follow.
Building Agents That Solve Real Operational Issues
The final day put every preceding lesson into action. In small groups, the Buna Constructions team designed and assembled AI agents targeted directly at operational pain points they identified themselves — from screening and evaluating candidate CVs against open roles, to classifying and routing employee feedback, to summarizing site and project updates for management review.
Each agent was built using the same disciplined approach introduced throughout the week: a clearly scoped knowledge base, explicit tools, and deterministic system instructions rather than free-form conversation. Trainers guided teams through testing, refining, and validating their agents against real inputs, ensuring what left the room on Day 4 wasn't a demo, but a working prototype ready to be handed to Buna's technical stakeholders for deployment.
The program closed with each group presenting their agent back to the full cohort — a moment that visibly shifted the room's confidence, as engineers and managers who had started the week unsure how to write a single prompt were now walking their colleagues through systems they had architected and built themselves.
"We came in thinking AI was something our IT department would eventually get to. We left with agents our own team built, already solving problems we've struggled with for months."
— Training Participant, Buna ConstructionsKey Outcomes from the Program
- ✓ Team-wide fluency in structured prompt engineering, applied to real Buna workflows
- ✓ Hands-on command of AI productivity and automation tools, including live Make.com workflow building
- ✓ A clear, shared mental model of AI agent architecture — LLM, memory, knowledge base, tools, and instructions
- ✓ Multiple working AI agent prototypes built in-house, targeting Buna's own operational challenges
- ✓ A trained internal cohort capable of extending and maintaining these systems going forward
- ✓ A repeatable framework for identifying future automation and agent opportunities across departments
Partner with dotpy.tech for Your Next Corporate AI Training
dotpy.tech designs and delivers practical, hands-on AI and automation training for organizations across the MENA region — from foundational prompting to production-ready AI agents built by your own team.
Start a Conversation