Completing the MIT Professional Education Chief Technology Officer Program on October 2, 2026 gave me an opportunity to connect my experience in enterprise data and AI with a broader perspective on technology leadership.

Working with Group 2 on MANAS brought that learning into practice. Our assignments developed a shared vision into a proposed architecture, R&D portfolio, AI strategy, and final pitch. Each stage raised different questions about feasibility, accountability, and the evidence needed to move forward.

Starting with the Problem

Our early work examined the barriers people face in recognizing distress, seeking help, accessing care, and maintaining support. Separating these challenges helped make the problem more specific.

My takeaway is to begin technology discussions with the people affected and the decisions that need to improve. That clarity helps determine where AI could contribute and where human relationships and established care pathways remain essential.

Turning a Vision into an Investment Plan

The roadmap and R&D portfolio required us to break the concept into workstreams with budgets, dependencies, and risks. Signal discovery, platform development, detection models, governance, and pilot measurement each needed attention.

This reinforced the importance of connecting investment to milestones. A roadmap becomes more useful when each stage identifies what must be demonstrated before the next commitment is made.

Distinguishing Prediction from Impact

The machine-learning assignment brought an important distinction into focus: identifying a pattern does not establish that acting on it improves an outcome.

Our proposed evaluation approach considered comparison groups and follow-up measurements. The final pitch presented improvements in recognition, reach, retention, and relapse as hypotheses for validation. I want to carry that discipline into future work by defining success clearly and separating promising ideas from demonstrated results.

Treating Responsible AI as an Architectural Requirement

The generative AI assignment expanded our evaluation beyond model capability. Offline operation, privacy, cost, safety controls, and human oversight influenced the proposed design.

The supervisor, screening layers, and clinician review process made accountability a concrete design consideration. For me, this reinforced that responsible AI requires explicit decisions about what a system may do, when it should defer, and who remains accountable.

Learning Through Collaboration

The group project required us to connect different professional perspectives in one coherent proposal. Technical detail had to support a story that others could understand, question, and evaluate.

The final pitch highlighted the importance of communicating assumptions and limitations alongside the opportunity. A credible proposal makes clear both what the team has developed and what still needs to be tested.

Applying the Learning Ahead

I intend to apply these lessons by defining intended use early, making assumptions visible, and linking technology investments to measurable evidence.

For MANAS, the next stage described in our final pitch is supervised feasibility work with appropriate partners. More broadly, this program has given me a framework for approaching AI initiatives with clearer priorities, stronger evaluation, and explicit accountability.