Converting caffeine into software...
Available for new opportunities · 2026

TwishaPatel

AI/ML Engineer · Software Engineer · LLM Systems

I build scalable software systems, healthcare technology solutions, AI-powered applications, and data-driven platforms that transform complex workflows into intuitive user experiences.

View Projects Download CV Contact
0
Projects Shipped
0
Years Experience
0
Open-Source PR Merged
0
BTech CGPA
twisha.py — engineer
1class Engineer:
2 def __init__(self):
3 self.name = "Twisha Patel"
4 self.role = "AI/ML Engineer"
5 self.stack = ["PyTorch", "LLMs",
6 "FastAPI", "Docker"]
7 
8 def build(self, idea):
9 # research → model → ship
10 return Product(idea).deploy()
11 
12Engineer().build("your idea")
AI LLMs · RAG
ML PyTorch · TF
FS FastAPI · Docker
About

Software Engineer at the intersection of healthcare, AI, and data.

I build the systems that ship — not just the models that demo.

I'm a Software Engineer building production systems at the intersection of AI, healthcare technology, and data engineering. As a Software Engineer at Artem HealthTech, I ship the integration software, clinical data pipelines, and Apache Superset dashboards that turn fragmented healthcare data into decisions clinicians and operators can act on. Earlier at Sundaram Technologies, I owned ML systems end-to-end — from computer vision pipelines for industrial inspection to FastAPI services engineered to survive real production traffic.

My focus is the engineering work that decides whether AI actually works in production: reproducible training, evaluation harnesses, latency budgets, and observability. The model is rarely the bottleneck — the system around it is. I'm currently sharpening that craft through an Advanced Post Graduate in Agentic AI Workflows at IIT Madras, turning research-grade ideas into dependable, production-ready software.

System Design
Scalable FastAPI services, REST APIs, data pipelines.
Healthcare Tech
Lab integration, HMIS reporting, clinical data exchange.
Data & Analytics
Apache Superset dashboards, BI, reporting on real datasets.
AI / ML Systems
LLM fine-tuning, RAG, computer vision, ML lifecycle.
"The model is rarely the bottleneck — the system around it is."
Experience

Where I've shipped.

Two years building production software across AI/ML, healthcare technology, and data platforms.

Software Engineer
Artem HealthTech · Ahmedabad, India
📅 Apr 2026 – Present 📍 On-Site
Current

Working on healthcare technology solutions focused on interoperability, data integration, and analytics.

  • Developing Lab Integration Software for healthcare systems — connecting laboratory systems to clinical applications.
  • Building and maintaining interfaces between laboratory systems and healthcare applications using REST APIs.
  • Improving healthcare data exchange and workflow automation across clinical and operational pipelines.
  • Working with clinical and operational data pipelines — ingestion, validation, transformation, and delivery.
  • Developing analytics and reporting solutions with Apache Superset for dashboards and insights.
  • Supporting HMIS Portal reporting and decision-making through clear, actionable data visualization.
  • Collaborating with product, healthcare, and engineering teams to deliver scalable, production-grade solutions.
Python FastAPI PostgreSQL Apache Superset Docker REST APIs Healthcare Integrations Data Analytics
Associate AI/ML Engineer
Sundaram Technologies · Vadodara, India
📅 Jul 2025 – March 2026 📍 On-Site
Industrial Automation
  • Architecting object detection and computer vision pipelines for industrial automation, replacing manual inspection workflows.
  • Building Python-based deep learning systems for image and video analysis on TensorFlow, PyTorch, and OpenCV.
  • Owning the full ML lifecycle — data preprocessing, model evaluation, deployment — to push detection accuracy on production data.
  • Partnering with product and ops to ship CV features that engineers can debug and operators can trust.
AI/ML Intern
Artem HealthTech · Ahmedabad, India
📅 Jan 2025 – Jul 2025 📍 On-Site
Healthcare AI
  • Built machine data integration solutions powering a healthcare platform's automation and patient-data flow.
  • Deployed real-time object detection using MMDetection's pre-trained COCO weights on live webcam feeds.
  • Developed agentic speech-to-text pipelines using open-source ASR models with multi-step reasoning.
  • Fine-tuned Flan-T5 on financial Q&A corpora, lifting domain accuracy through systematic prompt optimization.
Featured Work

AI products and platforms.

End-to-end case studies — from problem framing to deployed systems.

2026

Open-Source Contribution to LlamaIndex

Shipped a FastAPI-based RAG example to a major LLM framework — merged into main after maintainer review.

  • Challenge: Lower the barrier to running private, offline RAG without API keys.
  • Solution: End-to-end FastAPI service using LlamaIndex + local Ollama LLMs with document ingestion, vector indexing, and a query API.
  • Impact: Contribution merged into the main repository, serving as a reference for the community.
Python LlamaIndex FastAPI Ollama RAG Local LLMs
View Repository
2025

Pneumonia Detector AI

Explainable medical imaging classifier with GradCAM visual explanations and a retraining feedback loop.

  • EfficientNet backbone with GradCAM for clinician-trustworthy predictions on chest X-rays.
  • Human-in-the-loop feedback and batch retraining pipeline with quality assurance.
  • Fully Dockerized for reproducible deployment in clinical/research settings.
PyTorch EfficientNet GradCAM Streamlit Docker
View Code
2025

AutoML Pipeline Service

End-to-end ML platform: dataset profiling, model selection, versioning, and benchmarking through a web UI.

  • Streamlit + FastAPI for dynamic training, inference, and result visualization.
  • Automated dataset profiling and model benchmarking with performance logging.
  • Designed for ML teams to standardize experiments and ship models faster.
Python FastAPI Streamlit scikit-learn Docker
View Code
2025

Finance-domain Small LLM

Fine-tuned Flan-T5 on curated financial Q&A — shipped as a domain Q&A assistant.

  • Systematic prompt engineering lifted contextual accuracy on finance-specific queries.
  • Deployed for domain-specific Q&A inside a real workflow.
  • Built the data pipeline, training script, and inference service from scratch.
Python Hugging Face Flan-T5 Transformers
View Code
2025

Real-time Object Detection

Live webcam object detection using MMDetection & RTMDet-tiny with COCO pre-trained weights.

  • Rendered bounding boxes and class labels on streaming video with low latency.
  • Hands-on work with the MMDetection API for model loading and inference.
  • Foundation for the live-detection feature shipped at Artem HealthTech.
Python MMDetection OpenCV RTMDet
View Code
2023

Traffic Sign Recognition

CNN classifier for the GTSRB German traffic-sign dataset with advanced augmentation.

  • OpenCV preprocessing pipeline and data augmentation for model robustness.
  • Trained on TensorFlow + Keras with class-balanced sampling.
  • Foundation project that kicked off the deep-learning journey.
Python TensorFlow Keras OpenCV
View Code
2023

Desktop Voice Assistant

Python voice assistant for reminders, music, web search, and natural-language commands.

  • Speech recognition + TTS for seamless two-way interaction.
  • Enhanced natural-language understanding for smarter command parsing.
  • Early exploration of conversational UX and intent handling.
Python SpeechRecognition pyttsx3
View Code
Skills

How I build.

Core competencies and the tools I reach for daily.

Software Engineering

System Design & APIs

Designing, building, and operating scalable backend systems that survive production traffic. From REST APIs and async pipelines to authentication, observability, and data integrity.

Python FastAPI REST APIs System Design Scalable Services PostgreSQL Docker CI/CD
Healthcare Tech

Healthcare Integrations

Lab integration software, HMIS reporting, and clinical data exchange built for healthcare workflows.

Lab Integration HL7 HMIS Portal Clinical Pipelines
Analytics

Data & BI

Dashboards and reporting on real operational data.

Apache Superset Data Viz Reporting ETL
Databases

Data Layer

Modeling schemas, writing efficient queries, and shaping data for analysis and reporting.

PostgreSQL MySQL Pandas NumPy
AI / ML

LLM Systems

From research notebooks to LLM-powered applications.

PyTorch Hugging Face LangChain LlamaIndex RAG
Frontend

Web & UI

Building responsive, accessible interfaces that translate backend complexity into clean user experiences.

JavaScript React HTML5 CSS3
DevOps / Cloud

Ship & Run

Containerized, reproducible, observable.

Docker Git & GitHub Linux CI/CD
Stack

Tools I reach for.

Categorized by proficiency — expert, advanced, working knowledge.

Python FastAPI PyTorch TensorFlow PostgreSQL Apache Superset Docker REST APIs OpenCV Hugging Face Transformers LangChain LlamaIndex Healthcare Integrations MySQL JavaScript React Git & GitHub System Design Data Analytics TypeScript Node.js MMDetection Flan-T5 Streamlit scikit-learn Pandas NumPy SQL HTML5 / CSS3
Education

Academic foundation.

Continuous learning — formal degrees and ongoing specializations.

Dec 2025 – Jun 2026
Advanced Post Graduate
Agentic AI Workflows & Agentic System Development
Indian Institute of Technology, Madras
Ongoing Specialization · Agentic AI
Aug 2021 – May 2025
Bachelor of Technology
Computer Science Engineering
ITM (SLS) Baroda University
CGPA 8.6 4-year degree
2020 – 2021
Higher Secondary (XII)
Science Stream
Shri J.R. Shah Bright Eng. Medium School
Grade B2
2019 – 2020
Secondary (X)
Foundation
Bright Day School
Grade B1
Let's build something

Have an idea? Let's ship it.

Open to full-time Software Engineer and AI/ML roles, healthcare technology and data platform work, and research collaborations. The fastest way to reach me is email — I reply within 24 hours.

twishap534@gmail.com Download Resume
Email
twishap534@gmail.com
GitHub
@twishapatel12
LinkedIn
Twisha Patel
Phone
+91 81608 93358