Ishaan Choubey

full stack engineer

always tinkering. i ship ai native products, from multi agent pipelines to production grade systems built to scale.

email/github/linkedin/x
[00]About

hey, i'm ishaan. i build full-stack products with a focus on clean architecture, fast iteration, and real user impact.

currently building, breaking, and figuring things out. based in bengaluru.

Languages
TypeScriptJavaScriptPythonC++SQL
Frameworks
Next.jsReactNode.jsExpress.jsFastAPIFlaskLangChainScikit-learnPyTorchHuggingFace
Databases
PostgreSQLSupabaseMongoDB
Tools
VercelGitHubDockerGCP
[01]Experience
ThinklyLabs
Remote
Founding Engineer
  • Built authr ai from 0 to $2K+ MRR in 3 months.
  • Delivered 5+ client projects and MVPs
  • Architected a multi-agent content system by building a RAG system for long term context.
  • Automated LinkedIn outreach managing 50+ conversations monthly, cut manual prospecting by 60%.
  • Shipped a full B2B SaaS from scratch with 10+ paid users
  • Implemented RAG based assistant that cut document lookup time by 70% across 400+ page documents.
TypeScriptNext.jsNode.jsSupabase
Sept 2025 – Present
Full-time
Mythics
Bengaluru, IN
Software Engineer
  • Engineered an AI analytics agent using DeepSeek that processed raw company data into dashboards.
  • Designed a workflow to filter datasets before model ingestion, improving dashboard accuracy by ~30%.
  • Built AI agents using the OpenAI API to parse PDFs and interact with spreadsheets across 500+ documents.
TypeScriptPythonPandasDeepSeek API
Feb 2025 – Aug 2025
Internship
Novus
Mysore, IN
Software Engineer
  • Worked on a Django REST monolith powering a hyperlocal commerce platform.
  • Improved delivery fulfillment and transit flows by ~20%, resolving bugs across the order lifecycle.
  • Refactored cart creation and state transition logic, improving order lifecycle reliability.
PythonDjangoPostgreSQLDocker
Jul 2024 – Nov 2024
Internship
Ramaiah University
Bengaluru, IN
Research Intern
  • Annotated and preprocessed live infant video data from Ramaiah Hospital for pain/no-pain classification.
  • Trained an object detection model on neonatal facial expressions for real-time clinical monitoring.
  • Built a system to run the trained model on live video feeds for clinical use.
PythonPyTorchOpenCVYOLOv8
Apr 2024 – Jul 2024
Internship
Doodle Health
Mysore, IN
Research Intern
  • Designed data models and dashboards to visualize mental health indicators across patient cohorts.
  • Built an AI system to detect and classify mental health distress signals from user-generated content.
  • Analyzed patient data to surface trends and patterns across clinical segments.
PythonFastAPIPandasOpenAI API
Apr 2023 – Jun 2023
Internship
[02]Proof of Work
01
livegithub
MediMitra
doctor appointment platform — credit-based scheduling, role-based access for patients, doctors and admins, real-time video.
Next.jsNeonVonage
Appointments · Video calls · Role-based access
02
livegithub
Pretex
typographic ASCII playground with three modes — generative shapes, live camera to ASCII, and editorial text that reflows on canvas.
Next.jsTypeScript
ASCII art · Canvas · Typographic playground
03
livegithub
Droply
file storage app with drag and drop uploads, folder organization, starring, trash management, file previews, and bulk operations.
Next.jsSupabaseImageKit
File storage · Drag and drop · Bulk operations
04
github
Emotion Classifier
text-based emotion classifier using TF-IDF, traditional ML, and an LSTM. random forest ~85%, LSTM crossed 97% accuracy.
PythonKerasScikit-learn
NLP · LSTM · Emotion detection
[03]Education
Vidyavardhaka College of Engineering
Mysore, IN
B.E Computer Science (AI&ML)
CGPA: 8.4
2021 – 2025