Research Scientist · RISE Lab, IIT Delhi
Sankar Balasubramanian
Applied AI Researcher · Personalized AI Architect · Human–AI Interaction Researcher · Agentic AI Systems Builder · Computational Designer · Affective Computing Researcher · XR Developer · Algorithmic Artist · & above all, a Teacher
Applied AI & Computational Design Researcher building personalized AI that understands how people think, feel, and who they are. My interests are four-fold: the Human — mind, emotion, and personality; the Agents and architectures that meet them; the Interactions between them — personal, trustworthy, and responsible; and the Design that weaves them together — on the path to human-centred superintelligence.

Research Lens
Four lenses on human–AI interaction
Personalized AI starts with understanding people — and ends with designing how people and AI work together.
The Human
Understanding people
How people think, feel and differ — modelled well enough for AI to adapt to the person, not just the prompt.
- Cognition
- Emotion
- Personality
Agents & Architectures
Understanding the AI
Agentic systems, structured prompting and models that adapt to physiological and neural signals.
- Architecture
- Prompting
- Neuro-adaptation
The Interaction
Understanding the space between
Interactions that are personal, trustworthy and responsible — where people stay in charge and trust is earned claim by claim.
- Personalization
- Trust & calibration
- Responsibility & privacy
- Agency & co-creation
- Social & affective
Design
At every layer
Design as the thread through all three — from the person to the service, the agent and the system.
- Human
- Service
- Agent
- System
Toward human-centred superintelligence — AI that amplifies people.
Selected Work
A window into my work
Studies, tools and agentic systems — from doctoral research to apps built for real people.

Indie Research · Human–AI Trust
HARICHAN
Do people blindly trust AI? Learners absorbed confident falsehoods at every reliability level.
Explore
Indie Research · Personalized AI
The Personality Project
A 116-facet Atomic User Model that tells an assistant how you like your answers.
Explore
Doctoral Research · Agentic AI
MIDAS
Novelty on demand: progressive ideation with global and local novelty checks.
Explore.png)
Indie Research · Agentic UI Design
AURA
From lo-fi sketches to hi-fi UI and a deployable React codebase, with UX laws as constraints.
Explore
Doctoral Research · Gaze & Form
RETINA
Agentic form generation from implicit visual attention captured in VR.
Explore
Application · Biosignals
Synapse Studio
Learn, simulate, generate and stream ECG, EMG, EEG, EDA and more — in the browser.
Explore.png)
Doctoral Research · Multi-agent Design
FENSO
Progressive, designer-led form generation for complex products.
Explore
Indie Research · Human–AI Ideation
VITIS
From an everyday issue to rendered concepts — and the illusion of novelty.
Explore
By the numbers
A decade of building, teaching and publishing
- Peer-reviewed papers
- 34
- Patents & designs
- 11
- Research projects
- 35
- Apps & devices built
- 52
- Best paper awards
- 3
- Awards & honours
- 16
- Talks & videos
- 17
- Workshops & invited talks
- 7
Code, by the day
A year of commits, as a skyline
Every GitHub contribution of the past year, one bar per day. Bar heights follow a logarithmic scale. @sankar-mechengg
- Contributions
- 1,732
- Active days
- 189 / 366
- Longest streak
- 12 days
- Busiest day
- 291
Affiliations & memberships
- IIT Delhi
- IISc Bangalore
- RISE Lab
- ACM
- IEEE
- ACL
- The Design Society
- Design Research Society
- Association for Designers of India
- AMM
- EdTech Society
Let's collaborate