Northeastern University · Khoury College of Computer Sciences
Expected Dec 2027Master of Science in Computer Science · Boston, MA
Relevant coursework
Programming Design ParadigmsDatabase Management SystemsAlgorithmsML OpsNatural Language Processing
Open to Summer 2027 SWE/ML Internships & Spring 2027 / Fall 2027 Co-ops
Shubham Hadawle
MS in Computer Science · Software Engineer · AI/ML Engineer · Data Scientist
I'm aFullstack DeveloperSoftware Engineer
Engineering AI systems, scalable backends, and production-grade software at the intersection of research and product. Currently engineering GraphRAG pipelines, Transformer systems, and full-stack platforms at the intersection of research and product.
~/shubham — zsh — 80×24live
press ⌘K for command palette
01 · About
Engineer at the seam of AI research and production software.
I build at the intersection of applied ML, scalable backends, and product engineering — turning research-grade ideas (GraphRAG, transformer architectures, time-series forecasting, neural ODEs) into systems that ship, scale, and survive contact with real users.
About
Shubham Hadawle
Boston, MA
I'm a graduate Computer Science student at Northeastern University's Khoury College of Computer Sciences, focused on Software Engineering, Machine Learning, Data Science, and AI systems. My work spans the full spectrum: research prototypes published in Springer, GraphRAG retrieval pipelines for an enterprise bank, neural-ODE simulations of nuclear reactor dynamics at BARC, and shipping monetised Android apps with real users.
I move comfortably between ML engineering, distributed systems, and full-stack product work. I care about measurable impacts and engineering tastes.
The thread that ties everything together is curiosity. I want to keep building agentic LLM systems, software applications, and distributed ML infrastructure — and ship them like a founder would.
“The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore, all progress depends on the unreasonable man.”
4+
Years engineering
2
Research experiences
1,000+
App downloads
96.7%
Model accuracy
Currently exploring
- Agentic LLM systems & tool-use orchestration
- Production GraphRAG retrieval over enterprise knowledge graphs
- Distributed training & inference for transformer workloads
- Evaluation harnesses for grounded, faithful LLM outputs
Education
University of Mumbai
Jul 2020 – Jun 2024Bachelor of Engineering, Artificial Intelligence & Data Science · Mumbai, India
Relevant coursework
Machine LearningArtificial IntelligencePythonData StructuresObject Oriented Programming (Java)
02 · Stack
The toolchain I've shipped with.
Languages, frameworks, ML infra, and research tooling I use end-to-end — from prototype Jupyter cells to production deployments on AWS and GCP. Hover any pill to see its brand accent.
Languages
PythonJavaCC++JavaScriptTypeScriptRShell
AI / ML
PyTorchTensorFlowKerasScikit-LearnHugging FaceLangChainLangGraphTransformersLLMsRAGGraphRAGSBERTOpenCVspaCyNLTK
Backend
Node.jsExpress.jsFastAPIFlaskREST APIsSQLAlchemyPySparkPostman
Frontend
React.jsNext.jsHTMLCSSStreamlitAndroid StudioXMLVite
Databases
PostgreSQLMySQLMongoDBFirebaseNeo4jSQLNoSQL
Cloud / DevOps
AWS (EC2, S3, Elastic Beanstalk)Google Cloud PlatformDockerKubernetesGitHub ActionsCI/CD
Tools
GitGitHubMATLABPostmanJupyterVS Code
Research
Self-AttentionTransformersNeural ODEsAutoencoders / VAEsGANsTime-Series ForecastingFacebook Prophet
03 · Experience
Engineering across software development, research, and ML.
Four engineering roles spanning machine learning for retrieval, GraphRAG over knowledge graphs, neural-ODE physics simulations, and native Android product engineering.
04 · Projects
Production-grade engineering work.
Selected projects spanning AI/ML systems, published research, full-stack platforms, and shipped mobile titles — each with measurable outcomes.
Featured · Mobile
Where is my Duck? — Platformer Game
Cross-platform 2D platformer · 1,000+ installs · 4.7★ rating
Problem
Ship a polished, monetised 2D platformer to both mobile and desktop and validate a real distribution & ads stack end-to-end.
Solution
Designed, built, and published a GDevelop5 platformer to Google Play, itch.io and standalone Windows/Linux/macOS builds, with AdMob monetisation and CTR-tuned ad placements.
- Designed & published the title to Google Play and as Win/Linux/macOS desktop builds, sustaining a 4.7/5 rating across stores.
- Monetised the Android build via Google AdMob with strategically placed interstitial + banner units.
- A/B-tested ad frequency to drive a +15% click-through rate without harming retention.
GDevelop5AndroidGoogle AdMobA/B Testing
duck-platformer preview · v1
1,000+
Installs
4.7 / 5
Rating
+15%
CTR uplift
More Projects
Scroll down to drive the rail across.
scroll ↓ · pans →
Research
Text Summarization of Traditional Indian Languages
Transformer · Self-Attention · Low-Resource Indic NLP
PyTorchTransformersSelf-AttentionIndic NLP
0.62
BLEU
1,000+ articles
Dataset
IndicBART
Baseline beat
Full-Stack
MixMaster — Full-Stack Cocktail Database Platform
React · FastAPI · MySQL · Dockerised full-stack
React.jsFastAPIMySQLSQLAlchemyViteDocker
3NF
DB form
Offline fallback
Resilience
Dockerised
Container
AI / ML
Anomaly Detection in Multivariate Time-Series
Facebook Prophet · Unsupervised ML · AWS Elastic Beanstalk
PythonFacebook ProphetPandasAWS Elastic Beanstalk
96.7%
Accuracy
CPC · CPM
KPIs covered
AWS EB
Deploy
AI / ML
Question Paper Generation from Handwritten Notes
OCR + T5 Transformer · MCQ / Short-Answer / True-False
PythonT5TransformersGoogle Cloud VisionOCRPyTorch
< 2 min
Latency
3
Question types
OCR → T5
Pipeline
More work
Many more projects, code & experiments live on GitHub.
Source code, case studies, and side experiments — explore the full engineering portfolio at github.com/shubham-hadawle.
05 · Research
Published research in International journals.
Two Springer publications across time-series anomaly detection and transformer-based summarisation for low-resource Indic languages.
01
Springer · Lecture Notes in Networks and Systems·2023·DOI 10.1007/978-3-031-35644-5_16
Anomaly Detection in Multi-variate Time Series Data using Facebook Prophet
Time-SeriesProphetUnsupervisedAWS
02
Springer · Lecture Notes in Networks and Systems·2024·DOI 10.1007/978-981-97-9112-5_29
Text Summarization of Traditional Indian Languages using a Self-Attention Mechanism
NLPTransformersSelf-AttentionIndic Languages
06 · Engineering Activity
Building in the open.
Public engineering footprint — contribution heatmap, repository signals, and the kinds of problems I keep gravitating toward.
@shubham-hadawle
Contribution graph · last 12 months
less
more
Public repos
30+
Languages
Python · TS · Java
Focus
AI · Backend · Full-Stack
Open to
Collabs · Research
Activity signal
Recent commit cadencedaily
Primary languagesPython · TS · Java
Areas of focusLLMs · RAG · Backend
07 · Contact
Let's build something together.
Open to Summer 2027 SWE/ML internships, Spring/Fall 2027 co-ops, research collaborations, and engineering conversations. Drop a note and I'll respond within 24 hours.
Available · responding within 24h
Based in
Boston, MA · Open to relocation
Emailhadawle.s@northeastern.eduLinkedInin/shubham-hadawleGitHub@shubham-hadawleAndroid PortfolioGoogle Play developer profile
$ say --hello
recruiters — fastest reply via this form or LinkedIn.
collaborators — GitHub DMs work too.