2025 — PRESENT
Director, Data Science
Fidelity Investments
AI strategy and the development of data-driven solutions.
LEADERSHIP · RESEARCH · ENGINEERING
Data science leadership.
Research-led thinking.
Director, Data Science at Fidelity Investments
My work spans applied AI, natural language processing, and speech. I bring a research background to the design and development of intelligent systems.
01 / EXPERIENCE
From research to
applied leadership.
2022 — 2025
Senior Data Scientist
Fidelity Investments
Applied AI and data science in financial services.
2018 — 2022
PhD Researcher
Carnegie Mellon University
Natural language and speech processing.
Earlier research experience
Carnegie Mellon University · Research Assistant, 2016–2018
IIIT Hyderabad · Research Assistant, 2014–2016
02 / SELECTED PROJECTS
Ideas, built
into systems.
Independent personal projects exploring how AI systems gather context, act, and stay accountable.
PERSONAL AI
Judith ↗
A self-hosted assistant spanning voice, group chats, and a personal newsroom.
NEWS & KNOWLEDGESangraha ↗
More than a million source records, connected into stories and narratives that evolve with the evidence.
AGENT SECURITYJataayu ↗
Prompt-injection detection and privacy protection for LLM-backed agents.
03 / RESEARCH & WRITING
A foundation
in language.
Selected work in speech, multilingual language processing, and applied AI.
All publications on Google Scholar ↗2019 · SURVEY
A survey of code-switched speech and language processing
S. Sitaram, K. R. Chandu, S. K. Rallabandi, A. W. Black
2017 · INTERSPEECHOn Building Mixed Lingual Speech Synthesis Systems
S. K. Rallabandi, A. W. Black
ESSAYS & FIELD NOTES
Writing about systems, context, and AI.
On Rallabandi Radar ↗Engineering field notes ↗04 / TALKS
Talks
and panels.
Speaking on personal agents, speech technology, and applied AI — at industry conferences, academic venues, and panels.
2026 · INVITED TALK — AI ENGINEER WORLD'S FAIR, SAN FRANCISCO
Wearing the Agent: Engineering a Family-and-Friends Personal Agent, from Group Chats to Glasses
Selected from 2,000+ submissions. A 19-minute session on the Judith personal-agent system, from group chats to glasses. Recording on the AI Engineer channel.
2022 · INDUSTRY PANEL — AI SUMMIT AUSTINNavigating NLP: adoption, bias, and explainability
Panel with Google and Goldman Sachs on practical challenges of NLP in production. Coverage: AI Business.
2022 · CMU LANGUAGE TECHNOLOGIES INSTITUTEPhD thesis defense: De-Entanglement — A Framework towards building Ubiquitous speech technologies
Doctoral defense, January 13, 2022.
2020 · CMU LANGUAGE TECHNOLOGIES INSTITUTEThesis proposal: De-Entanglement — A Framework towards building ubiquitous speech technologies
Proposal talk, August 6, 2020.
2019 · SPHINX LUNCH TALK, CMU LTIOn Controlled De-Entanglement
Speech-group talk, Fall 2019. Accompanying writeup.
CONFERENCE PRESENTATIONS
Selected paper presentations.
Oral: "Variational Attention Using Articulatory Priors for Generating Code Mixed Speech" — Interspeech 2019, Graz. Oral: "On Building Mixed Lingual Speech Synthesis Systems" — Interspeech 2017, Stockholm. System presentation: "Submission from CMU for Blizzard Challenge 2019" — Vienna. Poster: "Acoustics Based Intent Recognition Using Discovered Phonetic Units for Low Resource Languages" — ICASSP 2021.
CONTACT