Case Studies

What we did

Multi-track engagement covering merger due diligence to turn NativelyAI into an integrated ecosystem for AI product development with an ongoing connection to a compute layer; product advisory; and dev-team AI enablement via Claude Code. Defined the AI-builder feature set to reach a competitive state. Onboarded the team to Claude Code, established a PRD-ingestion process, and defined product commercialization with competitive subscription tiers and a referral system. Additional features introduced: multi-agent infrastructure, a document layer for user projects, intelligent context management with file and message summarization, a planning agent, and parallel work with resources.

Outcome

Formation of Native.Compute as the decentralized AI compute layer. Developer velocity increased 8× on average, with top contributors hitting 12×. Commercially viable product launched.

What we did

Pre-implementation workshop to surface the main pain points in the architecture firm’s quality workflow. Defined the vision for a QAQC AI agent working directly against Revit models, providing instant quality assurance against both internal and external rules. The system comprises a Revit MCP layer, connector, agentic layer, a multi-agent work-parallelization system, and a building-codes-and-rules ingestion system.

Outcome

60% reduction in issues in the engineers’ work; 40% decrease in rework cycles.

What we did

Series of public AI workshops for business audiences. Agenda covered how AI works under the hood, advanced system-instruction design, principles of autonomous agentic AI-native companies, and concrete use cases and applications.

Outcome

Attendees went from understanding how AI works and how to steer it through to building first practical implementations live in-session. Businesses left with applied insights across marketing, sales, customer support, employee onboarding, maintenance, and production.

What we did

Delivered the AI acceleration experience for Brinc’s Middle East SDG 3 (health-focused) enterprise adoption program. The engagement spanned three workstreams:

• Workshop design — two stages, from problem-statement discovery and selection through to tech-fit identification and validation of proposed implementation paths.
• AI startups assessment — scored 38 AI ideas against viability, tech readiness, SDG 3 fit, adoption probability, and other criteria.
• Corporate AI playbooks — 70–100 page documents, interactive notebooks, and customGPTs guiding the corporation through gap identification, validation, and enterprise adoption of AI solutions.

Outcome

12 healthcare organizations completed the program; 38 AI ideas evaluated; 10 AI initiatives kicked off as intrapreneurship programs aimed at SDG 3 goals.

What we did

Interactive onsite workshop on agent setup for retail and shopping-centre management. Participants split into small working groups and worked on specialized AI use cases across operational areas — customer engagement, personalization, tenant management, employee onboarding, and more.

Outcome

Participants gained hands-on experience writing system instructions, testing agents, and rolling them out in their own operational contexts.

What we did

Multi-stage AI enablement program for a publicly-traded company in the KSA food sector. Programs were delivered separately to the CEO, vice presidents, directors, managers, and the IT & Digital Transformation department.

Outcome

AI enablement across the company — opportunity discovery, first AI prototypes, and agent implementations spanning finance, GRC, sales, marketing, operations, agriculture, and HR. The engagement also established a protocol for AI-enabled working groups and an opportunities-identification process using high-level BPMN notation. Prototypes leveraged AILER by ADAIA, n8n, Microsoft Copilot Studio, Microsoft Power Automate, and proprietary platforms.

What we did

Designed an AI roadmap as a transformation driver, repositioning Timeless Properties as an umbrella brand spanning luxury and wealth-preserving categories.

Outcome

1,000+ new leads generated post-engagement, plus strategic clarity on the company’s growth direction.

What we did

Built AILER, an agentic AI platform that lets enterprise clients spin up an agent in roughly 30 minutes. Capabilities include interactive forms, agentic personas, scheduled runs, and customer file generation. Used in multiple ADAIA Practice engagements as the underlying delivery vehicle.

Outcome

Enterprise clients spin up and test AI agents across departments in dozens of deployed cases — website-visitor engagement in consulting, shopping and retail, real estate, and reservations; internal assistants for corporate learning and knowledge management, project management in advertising, AEC, social media and content, and staff admin support.

What we did

Delivered a public workshop on AI-native startup setup — problem identification, team formation, strategy formulation, and product vision — plus hackathon-preparation guidance. Also served as judge on the AI hackathon at /function1 in Dubai.

Outcome

Educated and judged 300+ hackathon teams; provided detailed feedback and improvement roadmaps to 15+ participating teams.

What we did

Letzgro pioneered AI chatbots using IBM Watson’s tools (Speech-to-Text, Text-to-Speech, DialogFlow) alongside Caffe, OpenCV, and NLTK for the broader AI category — at a time when most of these primitives were still novel commercial offerings.

Outcome

Top-candidate status on Upwork and recognition as one of the early movers in the AI chatbots category.

What we did

Full agentic system for customer-support activities, designed for white-labelling across enterprise surfaces. The bots connected to CRM systems, a customer authentication layer, and arbitrary technical APIs for autonomous troubleshooting — effectively a prototype of what MCP would later standardize — plus calendar integration for appointment scheduling.

Outcome

Fully working solution shipped end-to-end, from initial design through implementation.

What we did

Built the first version of the SupportLogic dashboard, aggregating sentiment data from customer-service conversations.

Outcome

First prototype that led to the seed round. The company has raised $62M+ to date.

What we did

Built a part-time job-matching agent connecting HoReCa-industry job seekers with employers. The agent led the negotiation and clarification of candidate constraints — schedule and availability, preferred work locations, skills, and cultural match.

Outcome

Transformed the company from a recruiter-heavy, limited-ops model into a scalable tech startup capable of sustained growth.

What we did

Ian participated as one of two developers building the data recovery layer for Microsoft Exchange, after multiple prior teams had failed to crack it. The problem: MS Exchange had no documented or extractable format for emails stored in the database, so the team had to reverse-engineer the binary format from the hexadecimal files themselves.

Outcome

Three-tiered application — DLL, task queue layer, and UI — that read emails out of corrupt disk data. The company was sold to Dell six years later at an approximate valuation of $130M.