Skip to content
View AthiraSreenath's full-sized avatar

Block or report AthiraSreenath

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AthiraSreenath/README.md

Hi there 👋

I develop intelligent systems with a strong emphasis on practical impact and scalability. With experience spanning high-growth startups and multinational corporations, I have delivered AI solutions across NLP, Computer Vision, and Voice AI. I build efficient, domain-specific models tailored to business needs - using both Gen AI and traditional ML algorithms.

Model Building: Built and fine-tuned models for various domain-specific tasks

  • Full finetuning, LoRA, QLoRA
  • Text classification: ModernBert, Bert, DistilBert
  • Image classification: YOLO, ViT
  • Custom NER modeling
  • Full document classification

Multi-modal RAG systems: For virtual assistant, document QA

  • Document parsing, chunking
  • Semantic search, keyword search, FAISS+HNSW, vector DBs
  • Re-ranker, LLMs

Voice AI systems: For virtual assistants

  • Automatic Speech Recognition (ASR)
  • Speaker diarization using agglomerative clustering
  • Voice cloning (Coqui TTS, Tortoise TTS, and ElevenLabs)

Document understanding: For downstream tasks like RAG, classification etc

  • Text parsing
  • Image and diagram extraction
  • Text extraction from images ( OCR)

Curating High Quality Dataset: For training, evaluation and benchmarking models

  • Curation of high quality custom real-world data
  • High quality synthetic data generation using LLMs
  • Data labelling, augmentation using LLMs
  • Data wrangling, visualization and analysis

Responsible AI: For responsible and safe AI usage

  • Red teaming LLMs: Created frameworks and then built datasets to evaluate LLMs for bias (against protected classes)
  • AI Audit: Creating SOPs for clients to ensure their AI development meets legal, ethical, and regulatory compliance.
  • Working on PII detection and custom NER modeling

Tech:

Generative AI, LLMs NLP , Computer Vision, Voice AI RAG, Responsible AI, Data Analysis, Data Visualisation Python, SQL Git, Ray

Individual Contributor & Team Player:

Comfortable working as an individual collaborator, having developed multiple models for domain-specific enterprise use cases. Also thrive in collaborative environment - wisdom of crowds is always a strength.

Leadership Skills:

With proven leadership skills, I have successfully managed small teams, overseeing their daily tasks and facilitating KT. I have led development and production teams for an ahead-of-schedule DataStage version upgrade.

Medium

Kaggle

Pinned Loading

  1. Multi-Modal-Embeddings-Using-ImageBind-with-FAISS-for-Similarity-Search Multi-Modal-Embeddings-Using-ImageBind-with-FAISS-for-Similarity-Search Public

    Python 1

  2. Comparing-Image-to-Text-Models-BLIP-BLIP2-and-LLaVA Comparing-Image-to-Text-Models-BLIP-BLIP2-and-LLaVA Public

    Jupyter Notebook

  3. Web-Scraping-and-Sentiment-Analysis-using-NLP Web-Scraping-and-Sentiment-Analysis-using-NLP Public

    Jupyter Notebook

  4. DatabaseQueries DatabaseQueries Public

    Python

  5. treeclassification treeclassification Public

    Python