Systems thinking meets client success
— from AI research to real-world results.
Published AI/ML researcher (AAAI · Springer · IARIA) and Manager of Technical Customer Success. I turn complex AI/data problems into outcomes clients can act on — and stay accountable through delivery.
A little terminal, for your questions.
How to use: 1) Click the terminal to focus · 2) Type a command starting with / · 3) Press Enter ↵ (or the button) to run.
Use ↑/↓ for history · clear to reset the output.
Two practices, one habit — frame the whole system, then make the next step obvious.
I'm a data science and AI strategy professional in the New York City metro area. I came to tech from a business and finance background in India, then earned my MS in Data Science from Pace University's Seidenberg School (2024), where I worked as a Graduate Research Assistant in the AI Lab under Dr. Christelle Scharff — and where I now teach as an Adjunct Instructor (CIS 101).
I run two practices in parallel. On the technical side, I'm a published AI/ML researcher — NLP, LLMs, and generative models — with papers at AAAI (presented at Stanford), Springer/RIIFORUM (Best Paper Award, Ravello 2024), and IARIA AISyS (selected and invited for an extended journal version). On the client side, I'm Manager of Technical Customer Success at Data Engineer Academy, where I build personalized study plans and guide learners through Python, SQL, system design, and data modeling — and get named in student reviews for it.
One habit connects both: frame the whole system end-to-end — data → model → visualization → decision — then make the next step obvious.
- role
- Manager, Technical Customer Success — Data Engineer Academy
- also
- Adjunct Instructor — Pace University
- focus
- AI/ML Research · Data Engineering · Client Success
De-risk your AI/data initiative.
- ·Peer-reviewed AI credibility (AAAI, Springer, IARIA) plus a documented client-success track record.
- ·AI strategy and governance literacy — from national policy review to practical guardrails.
- ·Generative-AI and NLP build experience, not just slideware.
- ·Translates technical work for non-technical stakeholders.
An AI/ML researcher who ships and communicates.
- ·NLP/LLMs, generative AI (GANs/diffusion), fraud analytics, data engineering.
- ·3 publications, 2 best-paper honors.
- ·End-to-end ownership: data → model → visualization → decision.
- ·Mentorship and mentoring experience across programs and classrooms.
- Current[img]
- 2025[qr]
- 2023–2024[img]
- Earlier[txt]India
Strategy Consultant, Adofast · Business & Market Researcher, Qyuki · AI/ML Intern, Shoptaki
Data-driven financial and operational analysis, plus dashboards.
Peer-reviewed AI/ML work — NLP, LLMs, and generative models.
- ORCID 0009-0006-4913-2444 ↗
- DBLP dblp.org/pid/376/8386 ↗
- Google Scholar search: Om Gaikhe ↗
- 2024Springer / RIIFORUM
AI Strategies: A Review of Selected National and Intergovernmental Approaches↗
A comparative read of how countries and intergovernmental bodies are actually writing AI policy.
In: Research and Innovation Forum 2024, Springer. DOI 10.1007/978-3-031-78623-5_1
🏆 Best Paper Award — Ravello, Italy - 2024AAAI Symposium Series
Inclusion Ethics in AI: Use Cases in African Fashion↗
What inclusive AI looks like in practice when the training data and the culture are African fashion.
AAAI Symposium Series, Vol. 3 No. 1. DOI 10.1609/aaaiss.v3i1.31266
Presented at Stanford - 2024IARIA AISyS
Towards AI-Generated African Textile Patterns with StyleGAN and Stable Diffusion↗
Training GANs and diffusion models to generate authentic African textile patterns, measured with FID.
AISyS 2024, IARIA — Venice, Italy
Selected paper · invited for extended IARIA Journal version
AI research chatbot over 1.7M+ ML publications
Built with BERTopic for topic modeling and Llama 2 for generation, plus a Longformer-based fine-tuned summarization model trained on arXiv papers.
Featured builds — code on github.com/Om-Gaikhe
Healthcare Fraud Detection with Neo4j↗
- problem
- Fraud rings in healthcare claims hide in relationships, not in single rows.
- approach
- Modelled providers, patients, and claims as a graph and queried it in Neo4j.
- result
- A graph-database approach that surfaces suspicious clusters investigators can follow.
neo4j · cypher · pythonFinancial Payment Fraud Detection↗
- problem
- Bank payment fraud is rare, imbalanced, and expensive to miss.
- approach
- Trained and evaluated ML classifiers on transaction features.
- result
- A model that predicts fraudulent bank payments from transaction patterns.
python · scikit-learn · sqlAI-Generated African Textiles↗
- problem
- Generative models rarely represent African textile design well.
- approach
- Fine-tuned StyleGAN2-ADA, StyleGAN3, and Stable Diffusion on textile imagery.
- result
- FID-evaluated generations, best FID 22.44.
stylegan2-ada · stylegan3 · stable diffusion
Features, publications, and milestones — a running log.
- 2025[milestone]
Esports team inclusion at university level↗
Instagram — Announcement
- 2024[feature]
Coding His Path: A Student's Quest for AI Mastery↗
Pace University — Feature
- 2024[paper]
Inclusion Ethics in AI: Use Cases in African Fashion↗
AAAI Spring Symposium
- 2024[paper]
Towards AI-Generated African Textile Patterns with StyleGAN and Stable Diffusion↗
AISYS 2024
- 2024[paper]
AI Strategies: A Review of Selected National and Intergovernmental Approaches↗
Springer — Book Chapter
“I would like to thank Om Gaikhe for his dedication to my study plan and success of the program.”
“Very happy with Om as he keeps me motivated at this down time of my career!”
- MS, Data SciencePace University, Seidenberg School of Computer Science and Information Systems · 2024
- Undergraduate — business & financeIndia
- ·CITI Program — Certified Researcher (IPS; Social & Behavioral Research)
- ·Forage — Quantium Data Analytics
- ·Forage — JPMorgan Chase Quantitative Research
- ·IBM Data Science Tools
Let's talk.
Reach out about AI/data consulting, customer success, mentorship, or mentoring.