Engineer focused on backend systems, ML pipelines, and AI-driven product experiences.
Final year B.E. Computer Science student at MITE, Karnataka - graduating 2026.
I build systems end-to-end from database schema design and REST API architecture to deploying ML models in production web applications. My work spans full stack development, clinical AI tooling, voice-driven interfaces, and enterprise CRM platforms. I approach engineering with a product mindset: every technical decision should create measurable impact.
Open to: Software Engineer · Backend Developer · Full Stack Developer · AI/ML Engineer
| Layer | Technologies |
|---|---|
| Languages | Python · JavaScript · C |
| Frontend | React.js · HTML5 · CSS3 · Bootstrap |
| Backend | Node.js · Flask · FastAPI · Laravel · REST APIs |
| Databases | MongoDB · MongoDB Atlas · MySQL |
| AI / ML | Scikit-learn · PyTorch · TensorFlow · LangChain · OpenCV · Tesseract OCR · MediaPipe · NumPy · Pandas |
| Cloud & DevOps | AWS EC2 · S3 · Ubuntu · Linux · Git · GitHub · Railway · JWT Auth |
| APIs & Tools | OpenAI API · ElevenLabs API · Spotify API · VS Code |
| Domain | Proficiency | Details |
|---|---|---|
| Machine Learning | ████████░░ Advanced | Scikit-learn pipelines, SMOTE, Random Forest, model evaluation (Precision / Recall / F1) |
| Deep Learning | ███████░░░ Intermediate | PyTorch, TensorFlow, image preprocessing, convolutional architectures |
| NLP & Conversational AI | ███████░░░ Intermediate | LangChain, RAG, OpenAI API, prompt engineering, text generation |
| Computer Vision | ██████░░░░ Intermediate | OpenCV, MediaPipe, image segmentation, object detection pipelines |
| OCR & Document AI | ████████░░ Advanced | Tesseract OCR, unstructured PDF extraction, clinical report parsing |
| Voice / Speech AI | ████████░░ Advanced | Speech Recognition, ElevenLabs TTS, real-time voice pipelines |
◆ Client Pulse - Project & Delivery Intelligence Platform | Active | April 2026 – Present
An enterprise-grade project and billing management platform designed to streamline the full software delivery lifecycle - from employee timesheet tracking to automated invoice generation across multiple billing models.
| Attribute | Details |
|---|---|
| Stack | Node.js · MongoDB Atlas · JWT · Jest |
| Architecture | RESTful API · Role-based access control · Modular service layer |
| Performance | Sub-500ms API response times |
| Test Coverage | 70%+ with Jest |
| Security | JWT authentication · Admin / PM / Employee / Client RBAC |
| Billing Models | Fixed Cost · Retainer-based |
| Impact | Eliminated manual invoice processing; unified project, timesheet, and billing data |
| Repository |  |
Architected and implemented the full backend from schema design through deployment. Built automated workflows that link approved timesheets directly to invoice generation, significantly reducing operational overhead. Designed MongoDB schemas with flexibility for multiple billing models, supporting diverse client engagement types on a single unified platform.
◆ Peripheral Arterial Disease Detection - Clinical ML System | Completed | March – November 2025
An end-to-end ML pipeline for detecting Peripheral Arterial Disease (PAD) from unstructured clinical Doppler reports, integrating OCR-based text extraction with statistical classification for automated diagnostic support.
| Attribute | Details |
|---|---|
| Stack | Python · Flask · Scikit-learn · Tesseract OCR · SMOTE |
| Model | Random Forest classifier with class imbalance correction |
| Accuracy | 75% model accuracy; improved minority class recall via SMOTE |
| Scale | Processes unstructured free-text clinical reports end-to-end |
| Impact | Automates clinical report triage; outputs clinically interpretable predictions |
| Team | 4 engineers |
| Repository | |
Designed and owned the ML pipeline: OCR extraction via Tesseract, feature engineering from noisy medical text, model training with SMOTE to address severe class imbalance, and a rule-based validation layer for clinical safety. Deployed as a Flask web application with real-time report upload, automated prediction, and human-readable output.
◆ ZOE - AI Voice Assistant | Completed | February 2025
An intelligent, real-time voice assistant built on Python and Flask, powered by OpenAI for language understanding and ElevenLabs for expressive voice synthesis.
| Attribute | Details |
|---|---|
| Stack | Python · Flask · OpenAI API · ElevenLabs API · SpeechRecognition |
| Modality | Voice-in / Voice-out with natural language understanding |
| TTS Engine | ElevenLabs - high-fidelity expressive voice synthesis |
| LLM | OpenAI GPT for context-aware conversational responses |
| Impact | Fully functional personal AI assistant with natural interaction loop |
| Repository | |
Built a complete real-time speech processing loop: microphone input → speech-to-text → LLM inference → expressive TTS output. The system handles open-domain queries, commands, and conversational turns with low latency through a lightweight Flask backend.
◆ Voice-Controlled Music Player | Completed | November 2025 – January 2026
A hands-free music control system that bridges natural speech commands with the Spotify API, enabling full playback control without a screen or manual input.
| Attribute | Details |
|---|---|
| Stack | Python · Flask · Spotify API · SpeechRecognition |
| Controls | Play · Pause · Track navigation · Volume adjustment |
| Architecture | Modular voice command processor with low-latency response pipeline |
| Feedback | Natural TTS confirmation for all executed commands |
| Impact | Fully hands-free Spotify control via conversational voice commands |
| Repository | |
Designed a modular backend to cleanly separate the speech recognition layer, command parsing logic, and Spotify API integration. Implemented audio feedback via TTS to close the interaction loop without requiring visual output.
January 2026 - May 2026
Worked on enterprise web applications, backend development, CRM systems, website redesign projects, and business automation solutions using Laravel and MySQL.
- Developed backend modules and business logic using Laravel
- Redesigned and performance-optimized client-facing websites
- Integrated REST APIs and relational database schemas
- Implemented secure authentication and role-based access control
- Delivered a full CRM/ERP-style Customer Support System with dashboards, CRUD, reporting, analytics, and notification modules
- Performed testing, debugging, and production deployment support
| Recognition | Details |
|---|---|
| 🏆 Prompt Wars | Winner - MITE Prompt Engineering Competition |
| 🎓 GPA 7.92 / 10.0 | Top academic performance, B.E. Computer Science, MITE |
| 🔬 Clinical AI Research | PAD detection system end-to-end ML pipeline in healthcare domain |
| 🚀 Production Platform | Client Pulse - live multi-role SaaS platform with billing automation |
| 🏢 Industry Internship | Software Engineer Intern at Udupi Web Solutions - enterprise CRM delivery |
Deloitte
MITE
ishita_acharya:
learning:
- System design and distributed systems architecture
- Advanced LangChain RAG pipelines and vector databases
- AWS Solutions Architect fundamentals
building:
- Client Pulse - delivery intelligence platform (active)
- Experimenting with multi-agent AI workflows
exploring:
- LLM fine-tuning and prompt optimization
- FastAPI microservices with async patterns
- MLOps and model deployment at scale
open_to:
- Software Engineer
- Backend Developer
- Full Stack Developer
- AI / ML Engineer
"Precision in engineering. Clarity in thinking. Impact in everything built."