Welcome to the Estate of

Jenica

JENICA A.

Reinforcement Learning · Graph Learning · Explainable AI

Welcome to my estate, dear reader, do come in! ✦

“Dear reader, this author is delighted to present a young lady of rare ambition. She studies how machines learn from consequence, builds systems that real people use, and presides over her own society besides. Do read on.”

Enter the Library Tour the Gallery Request my Card

8.72CGPA / 10

2Apps Live in Production

3IEEE Society Roles

2028Graduating Cohort

Room the First

The Study

Dear reader, allow this author a proper introduction. Miss Jenica is a third-year student of Artificial Intelligence & Machine Learning at Saveetha School of Engineering (SIMATS) in Chennai, to be presented to the world in May of 2028. She builds her systems from end to end and sends them out to live in the world, rather than leaving them to gather dust in a notebook.

Her scholarly affections lie with credit assignment in reinforcement learning, graph neural networks, and fairness in model explanations. Her practical talents run to predictive modelling and patient feature engineering. She means, in time, to make research her vocation.

In society she keeps a remarkably full diary: Chair of IEEE Women in Engineering, Membership Coordinator of her IEEE Student Branch, and Webmaster of the IEEE Computational Intelligence Society chapter.

DegreeB.Tech, AI & Machine Learning

HouseSaveetha School of Engineering, SIMATS

StandingCGPA 8.72 / 10

GraduationExpected May 2028

ResearchRL Credit Assignment · GNNs · XAI

Engagements

Machine Learning RoleSales Forecasting with LSTMs

Built an LSTM model forecasting sales across 50 product categories, reaching 12.4% MAPE against a 16.8% baseline. Follow-up analysis found 4% higher sales for women’s products, which informed inventory allocation and seasonal marketing.

IncomingData Engineering Intern · HCL

Upcoming internship in data engineering.

Room the Second

The Library

The young lady’s research, dear reader, her most learned pursuits.

Classical Graph-Algorithm Execution as a Controlled Testbed for Credit Assignment in Reinforcement Learning

Solo paper · Preprint, arXiv submission in progress (cs.LG)

Credit assignment, deciding which earlier action deserves the reward that arrives later, is among the oldest puzzles in reinforcement learning. This paper uses the step-by-step execution of classical graph algorithms such as BFS, DFS and Dijkstra as a controlled setting in which to study it, pairing graph neural networks with reinforcement learning.

Reinforcement LearningCredit AssignmentGNNAlgorithmic Reasoning

Bias-Driven Efficiency in AI using GNNs

Four posters · Tech Star Summit 2026

A single notion pursued down four separate avenues, pairing graph neural networks with PPO, DQN, GCN and Node2Vec to ask how bias might be turned toward efficiency. All four were presented together at the Star Summit.

PPODQNGCNNode2Vec

4 posters · presented in one sitting

Room the Third

The Gallery

A collection of the young lady’s finest works, dear reader, each framed and hung with care.

Adrian AI

RAG-Powered Study Companion · Deployed

A clever companion, dear reader, that reads a lady’s documents and answers only from what they truly contain, built upon her own pipeline for chunking, tokenisation and retrieval, with no borrowed framework to lean upon. Should one model falter, a chain of five attends in its place, so the conversation never once stumbles.

RAGLlama 3.2OpenRouterStreamlit Cloud

Live in production · zero infra cost

Orbit

Cross-College Collaboration Platform · Deployed

A drawing room for student builders across many colleges, raised entirely by her own hand. Within one finds secure entry by passcode, a lively feed of reactions, an AI matchmaker for collaborators, and a dashboard for keeping good order. Quite the accomplishment for one application.

JavaScriptAnthropic APIPWADesign System

Auth · Feed · AI · Admin, in one file

Predicting F1 Pit Stops

Kaggle Competition · Ensemble

A spirited wager upon the racing season, dear reader. The young lady forecasts the pit stops by reading the smallest of signs: the wear of a tyre, the drift of lap times, the progress of the race. Three fine models were trained and blended together with the greatest care.

LightGBMXGBoostCatBoostFeature Eng.

AUC 0.948 · 5-fold CV ensemble

Predicting Irrigation Need

Kaggle Competition · ML Pipeline

A thorough study across some ten thousand samples, by which she raised her accuracy fifteen per cent above her first baseline. More than twenty new features were fashioned by hand, and the labour of experiment cut by near a third. Industry, dear reader, of the most admirable sort.

Scikit-learnXGBoostEDA

+15% accuracy over baseline · 20+ features

Room the Fourth

The Conservatory

Her instruments, her diplomas, and the many distinctions she has gathered, dear reader.

Instruments of the Trade

PythonSQLJavaScript / HTML / CSS Scikit-learnXGBoost / LightGBM / CatBoostFeature Engineering RAG & LLM IntegrationReinforcement Learning (PPO, DQN)Graph Neural Networks LSTM ForecastingPandas / NumPyMatplotlib StreamlitOpenRouter & Anthropic APIsGit / GitHub

Leadership & Awards

Chair · IEEE Women in Engineering Student Chapter

Leads technical meetings, expert seminars and community events, including InnovateHER’26 for IEEE Day

Membership Coordinator · IEEE Student Branch, SIMATS

Appointed October 2026 · IEEE Star Recruiter (Bronze tier)

Webmaster · IEEE Computational Intelligence Society Student Chapter

Manages the chapter website and digital outreach for technical events

International Achievers Award · Silamboli Foundations (SIMATS)

Recognised for consistent performance in academics, leadership and student initiatives

Member · IEEE, IEEE WIE & IEEE Computer Society

Active member across three IEEE communities

Diplomas & Distinctions

Stanford Machine Learning Specialization

Foundations of supervised & unsupervised learning

Google Data Analytics · Professional Certificate

Data cleaning, analysis & visualisation

Google Project Management · Certificate

Planning, execution & delivery

Python Certification · Codecademy

Programming fundamentals & applied Python

Forage Job Simulations · JPMorgan Chase, Amazon, British Airways

Virtual programmes across finance, technology and operations

A Secret Room

The Parlour

You have found the hidden room, dear reader. Here the young lady sets her work aside a while.

From Her Pen · Essays

On Being a Woman Who Builds

Notes from a seat that wasn’t always offered.

I build things: models, apps, and small systems that take an input and hand back something useful. For a long time I assumed I had to ask permission to call myself a builder, as though the title belonged to someone louder, older, or more certain than me. It doesn’t. The first time I shipped something that worked for a person other than myself, I understood the rule: you become a builder by building, not by being told you are allowed to.

Being a woman in this field means I am often the only one in the room who looks like me. I used to shrink at that. Now I read it as information, proof the room still needs more of us, and that I have every right to take up space in it. So I keep things simple: do the work, document it honestly, and let it speak. The work has never once asked my gender before it ran.

Jenica A.

What the Machines Are Teaching Me

On AI, and the questions behind the metrics.

People expect AI to be about machines. Mostly, it has taught me about myself. A model is a brutally honest mirror: it learns exactly what you show it, biases and shortcuts included, and then repeats them back at scale. You cannot bluff a confusion matrix.

Machine learning has made me more patient and more skeptical at once. Patient, because the interesting results live on the far side of a hundred unglamorous experiments. Skeptical, because a number that looks impressive is usually hiding a question I forgot to ask. I’ve learned to chase the question, not the metric, and that the most powerful thing I can give a system isn’t more data or more compute, but a clearer idea of what “good” actually means. That, it turns out, is the hardest part of any problem, machine or human.

Jenica A.

Room Enough for All of Us

A few thoughts on women, and the table we sit at.

There’s an old idea that opportunity is a fixed pie: that for one woman to rise, another must fall, and that we ought to compete for the single seat we’re offered. I don’t believe it. Every time a woman builds something, teaches something, or simply stays in a field that wasn’t designed for her, the table gets a little longer.

I think often about the women who will come into engineering and AI after me, and about making the path a little less lonely than it can feel. That doesn’t take grand gestures. It looks like answering a beginner’s question without making her feel small, sharing what I learned the hard way, and refusing to be the kind of “only one” who pulls the ladder up behind her. The future I want isn’t one where women have finally proven they belong. It’s one where no one thought to ask.

Jenica A.

The Reading Chair

Classics & quiet evenings

A most devoted reader, dear friend, happiest with a classic in hand: the sort of novel where wit, longing and a well-stocked library all signify. It is little wonder a Regency estate should feel so like home to her.

Classic LiteratureSlow Reading

The Music Box

On constant rotation

When the young lady is at work, the music is unapologetically K-pop. BTS for the ambition and the feeling of it, BLACKPINK for sheer momentum. Good engineering, after all, deserves a good rhythm.

BTSBLACKPINKK-Pop

Room the Fifth

The Season’s Calendar

The young lady’s diary of the seasons, dear reader, year by year.

2024

Foundations

She completed the Stanford Machine Learning Specialization and the Google Data Analytics Professional Certificate, and so began to build in earnest.

2025

Taking the Lead

The young lady took the chair of the IEEE Women in Engineering chapter, and the post of Webmaster at the IEEE Computational Intelligence Society chapter.

2026

The Building Year

A most industrious year. She sent Adrian AI and Orbit out to live, presented four posters at the Tech Star Summit, wrote her first solo research paper on credit assignment in reinforcement learning, and was honoured with the International Achievers Award. In October she was appointed Membership Coordinator of her IEEE Student Branch.

2028

Commencement

The young lady is to be presented to the world with a B.Tech in Artificial Intelligence and Machine Learning from Saveetha School of Engineering, SIMATS.

Room the Last

The Correspondence

Should you wish to begin a correspondence, dear reader, whether upon research, a collaboration, or a position, the young lady is most warmly disposed to receive your letter.

Email LinkedIn GitHub Orbit · Live Adrian AI · Live Download my Résumé

Chennai, India

~ J.A.