Aalok Thakkar: Research

N.S. Bendre, Cosmos in the Making (1963)

My research focuses on logic in computer scienceThe study of formal systems for reasoning about truth, inference, and computation, and its applications to programming languages,The design, semantics, and implementation of languages used to express computational processes. automata theory, The study of abstract models of computation and formal languages. and trustworthy AI.Methods for ensuring that AI systems are reliable, fair, transparent, and aligned with human intent. Recently, I have also been exploring interdisciplinary questions in computational sustainabilityThe application of computational methods to problems in environmental and social sustainability. and CS pedagogy.

Formal Methods and Logic

My doctoral research focussed on automated program synthesis. A good starting point to learn about this work is this survey. In program synthesis, I am looking at the following two questions:

  1. Given a set of specifications, how do we find a program that satisfies a maximal subset of the specifications?
  2. Given a database transformation, how can we generate a query on the transformed database that is equivalent to a specified query on the original database? This project is in collaboration with TCS Research.
Beyond program synthesis, I am interested in various aspects of automata theory, logic, formal methods, and logic and probablistic programming.

Group:

  1. Smayan Agarwal (undergraduate thesis): Finite State Representations for String Samplers
  2. Shrey Arora (undergraduate thesis): Query Regeneration under Schema/Value Transformations
  3. Kudakwashe Chakanyuka (undergraduate thesis): Interfaces for Security Protocol Verification
  4. Arnab Ray (visiting researcher): Invariants for Probablistic Programs
  5. Tanush Seal (graduate student): Approximate Inference for Probablistic Logic Programs

Trustworthy AI

I am exploring several aspects of trustworthy AI:

  1. In collaboration with Moolya, we are developing methods to evaluate real-world AI systems, identify failure modes, and ensure reliable behavior under diverse scenarios.
  2. AI Safety for Bharat: Supported by the EkStep Foundation’s AI and Data Adoption Fellowship, this project aims to create context-appropriate safety benchmarks tailored to the Indian ecosystem. This work is conducted jointly with Debayan Gupta and Anirban Sen.
  3. We are investigating how enforcing Lipschitz continuity in transformer architectures can improve robustness, stability, and certifiability of large language models under adversarial or distribution-shifted inputs.

Group:

  1. Pranav Aggarwal & Aishwarya Pokhriyal (research assistants): AI Safety for Bharat
  2. Kashyap J. & Ananya Basotia (undergraduate thesis): AI Safety for Bharat
  3. Pranav Jayanandan & Prabhpreet Setia (undergraduate thesis): Lipschitz Continuous Vision Transformers

CS Pedagogy

I am interested in the impact of AI on computer science education and in developing curricula and tools that are relevant to the modern learning environment. This includes:

  1. Designing courses that emphasize conceptual understanding, formal methods, and systems thinking to go beyond what AI tools can solve automatically. We have built the BOOP framework for this.
  2. Developing intelligent assistants and teaching tools that guide learning while preserving active engagement and critical thinking.
  3. Creating educational resources and platforms that are accessible to all students, including those with disabilities. More details here.

Group:

  1. Vaani Goenka (research assistant): BOOP Framework
  2. Ayush Thonge (research assistant): AI assistants for teaching
These projects are supported by Mphasis Limited as a part of the Mphasis AI & Applied Tech Lab at Ashoka.

Computational Sustainability

I am working on two projects that bring together computation and the environment:

  1. Darjeeling Revival Project: We are developing statistical models to estimate carbon captured by Enhanced Rock Weathering (ERW) in Northern Bengal. This project is funded by Alt Carbon.
  2. OpenGIS India: This is a project in collaboration with Meghna Agarwala , where we use GIS and remote sensing to study forest dynamics, changing land usage and forest fire patterns, floods and river dynamics, and animal-human conflicts across India.

Group:

  1. Devesh Bajaj & Krish Goenka (undergraduate thesis): Predictive Models for Tea Yield
  2. Rajdeep Das, Karan Kumar, & Vedant Gautam (research assistants): Statistical Models for Enhanced Rock Weathering
  3. Taha Ibrahim Siddiqi (research associate): Remote Sensing for Inundation Mapping
  4. Kabir Vohra (research assistant): OpenGIS India

Former Members of the Research Group:

  1. Roshni Agarwal, Darjeeling Revival Project 2025 (Ashoka Class of 2026)
  2. Suyog Joshi, AI Safety for Bharat 2025 (Machine Learning Analyst, Wadhwani AI)
  3. Pankhi Mehta, AI Safety for Bharat 2025 (Ashoka Class of 2026)
  4. Shobhit Singh, Pre-doctoral Researcher (PhD Student, Karlsruhe Institute of Technology)
  5. Aaryan Shah, AI Safety for Bharat 2025 (Data Scientist, Fractal AI)
  6. Karthik Sunil, CS Pedagogy (Ashoka Class of 2026)