01The challenge
Older adults are often an afterthought in AI product design. The project asked how a conversational agent could support emotional well-being and meaning-making late in life, not just answer questions.
02What I did
- 01
Ground the design in theory
Drew on Erikson’s concept of ego integrity and Validation Therapy to define what a good conversation should do for this audience.
- 02
Design a three-stage dialogue
Structured conversations as Validation, then Meaning, then Legacy, moving from emotional acceptance to meaning-making to recording personal memories.
- 03
Test with real users
Ran sessions with 30+ older adults and synthesized what worked, what confused people, and where the AI missed emotional cues.
- 04
Propose how to measure it
Wrote a research proposal pairing psychometric scales with linguistic sentiment analysis so outcomes could be evaluated, not assumed.
