AI Engineer · Computational Bioengineer

Hello, I am Ruman

I'm currently studying Computational Bioengineering at Imperial College London.

Engineer

Researcher

Writer

Experience

AI Engineer, IBM

Client Engineering · Bangalore, India

Oct 2023 - Jul 2025

Built Retrieval-Augmented Generation (RAG) systems with 75-85% average retrieval accuracy across evaluated use cases.
Fine-tuned LLMs using synthetic data, improving response alignment by 25-35% based on expert evaluation.
Engineered MLOps pipelines with 90% drift detection accuracy and automated retraining.
Worked on RAG variants, agentic systems, and fine-tuning to reduce POC iteration time.
Developed benchmarking metrics for LLM evaluation across sectors.
Stack: Python, PyTorch, WatsonX, MLOps.

Software Engineer, Oracle Health (formerly Cerner)

Population Health Analytics · Bangalore, India

Jun 2021 - Aug 2023

Healthcare Data Analysis: Developed and tested pipelines to evaluate patient records for provider scoring systems.
Performance Optimization: Reduced pipeline runtime by 90%, shrinking report turnaround from multiple days to under one.
ETL & Automation: Used Apache Airflow and Spark to automate client workflows and speed up reporting.
Software Development: Delivered production-ready systems using Python, Oracle SQL, Java, Git, and Jira.

Research

Dissertation Research, Imperial College London

MSc Computational Bioengineering · Ongoing

Feb 2026 - Present

Investigating whether generalised operator-learning models can match or outperform dataset-specific physics-informed models in modelling microbial ecosystem dynamics governed by the generalised Lotka-Volterra system.
Developing and comparing PI-DION, baseline PINNs, and exploratory QPINN variants for parameter inference and trajectory modelling under data-scarce conditions.
Using biologically constrained synthetic data and real microbial time-series datasets to evaluate generalisation, sample efficiency, predictive accuracy, parameter recovery, and training cost.
Building a reproducible experimental pipeline with PyTorch, Hydra, Weights & Biases, Optuna, Ray, Docker, and PennyLane for model development, tracking, and evaluation.

Impact Scholar, NeuroMatch Academy

Impact Scholar Program - Virtual, Global

September 2024 - April 2025

NeuroAI Research: Investigating the role of architectural priors in biological and artificial agents for task performance.
Comparative Analysis: Modeled C. elegans (swimming) and rat (running) neural architectures, comparing them with MLPs to analyze deviations from parameter scaling laws.
Imitation Learning & RL: Applied teacher-student distillation and reinforcement learning techniques to establish empirical scaling curves.
Findings: Demonstrated that standard MLP scaling does not align with biologically evolved architectures.

Education

Imperial College London, London, U.K.

Masters of Science in Biomedical Engineering

2025 - Present

Specialisation in Computational Bioengineering.
Planned focus on AI-driven simulation of biological systems and neuro-inspired computing.
Selected modules: Brain Machine Interface, AI for Drug Discovery, Biomematics, Digital Biosignal Processing.

NHCE, Bangalore

B.E. in Computer Science

2017 - 2021

Graduated with distinction.
Focused on AI and systems programming.
Worked on academic and hackathon projects.

Focus Areas

A few things I care about and keep coming back to.

Computational Engineering

Bridging AI, neuroscience, and biology through code to build for real-world health impact.

AI for Science

Leveraging machine learning to accelerate discovery in health and life sciences.

Systems Thinker

Keeping the big picture in view across biology, tech, and society to guide meaningful design.

Projects

Prediction of Future Continuous Motion States from ECoG Recordings

NeuroMatch Academy , July 2023

sciKit-learn, pandas, matplotlib, Research

Built a data pipeline to analyze ECoG data for correlations between neural signals and cursor movement.
Achieved a max R-square of 49.3%, with processing latency reduced to 5 milliseconds using techniques like frequency filtering and PCA.
Identified correlations between Brodmann areas and neural signals, enabling faster processing by targeting specific brain regions.

Adversarial Tweet Sentiment Analysis

NeuroMatch Academy , June 2022

PyTorch, HuggingFace, Transformers

Performed sentiment analysis using SBERT from HuggingFace, reducing high-dimensional data to 3D space with PCA.
Trained a logistic regression model, validating data compression without loss in prediction accuracy.
Highlighted classification challenges with slang and Twitter-specific words, identifying model limitations.

Autonomous Ice Hockey Agent

UT Austin (Virtual) , May 2023

Python, PyTorch

Developed an autonomous agent to play ice hockey using image-based and state-based approaches.
Achieved 85% accuracy in ball tracking and over 80% game success through policy optimization with the REINFORCE algorithm.
Built a data pipeline for training and assembling datasets to optimize the agent's goal-scoring strategy.

Analyzing Dataset Artifacts using ELECTRA

UT Austin (Virtual) , June 2023

PyTorch, HuggingFace, Transformers

Developed an NLI model using ELECTRA, achieving 88.24% accuracy and improving predictions by correcting dataset artifacts.
Designed an error analysis framework, categorizing issues to enhance semantic processing and model robustness.
Conducted experimental fine-tuning on diverse datasets, improving model generalizability.

Publications

Peer-reviewed and preprint work I've contributed to.

View All

Toolkit

The areas I work across, day to day.

Machine Learning
Data Engineering
Research
Scientific Writing
Systems Thinking
Full-Stack Engineering