A

Aaweg Bhaladhare

About

Detail

Pune, Maharashtra, India
Contact Aaweg regarding: 
work
Full-time jobs
Starting at USD5k/month
Flexible work
Starting at USD30/hour
groups
Networking

Timeline

work
Job
school
Education
folder
Project

Résumé

Jobs verified_user 0% verified
  • T
    Team Lead
    Turing Enterprises Inc.
    Jan 2025 - Current (1 year 7 months)
    • Worked extensively on SFT and RLHF notebooks to improve client LLM performance across multiple benchmarks, focusing on correctness, reasoning quality, and instruction-following. • Handled multi-domain model evaluation and training tasks including Python (competitive programming), JavaScript, C/C++, and Mathematics, ensuring high-quality data and consistent outputs. • Served as Team Lead during Terminal Bench and OS Bench projects, managing task allocation, resolving blockers, reviewing submissions, and maintaining quality standards under tight timelines. • Applied system design principles to structure complex LLM evaluation workflows and validation rules for Terminal Bench. • Utilized Java knowledge to contribute to the development of spe
  • T
    ML intern
    Tars Technologies
    Jun 2024 - Dec 2025 (1 year 7 months)
    • Developed an object detection model using YOLOv8 to enhance safety monitoring, achieving 95% accuracy in detecting protective gear for mine workers. • Collaborated on a full-stack LMS using Next.js 13, React, Prisma, Tailwind, and MySQL, improving user experience and streamlining workflows. • Built a face recognition-based attendance system, automating tracking for 50+ employees and reducing manual errors by 30%. • Mentored junior engineers on best practices in containerization, model integration, and code reviews. • Applied system design principles, Java programming, CI/CD pipelines, distributed systems concepts, and cloud architecture knowledge to architect the full-stack LMS and containerize ML models for efficient deployment and manag
Education verified_user 0% verified
  • Vishwakarma Institute of Information Technology
    B.Tech in AI&DS
    Vishwakarma Institute of Information Technology
    Jan 2021 - Current (5 years 7 months)
    CGPA 8.54
Projects (professional or personal) verified_user 0% verified
  • N
    NibbleCore: LLM Benchmarking & Inference Engine (In Development)
    Jan 2026 - Apr 2026 (4 months)
    • Architecting a custom LLM evaluation framework designed to execute complex, OS-level benchmarks within secure, automated Docker-in-Docker sandboxes, leveraging cloud architecture principles to ensure robust and scalable sandbox environments. • Implementing low-level C++ quantization kernels using ARM NEON SIMD intrinsics, targeting maximum dequantization throughput for Apple Silicon over standard llama.cpp baselines. • Designing recursive multi-agent evaluators to autonomously test and score LLM reasoning logic and instruction-following within dynamic system environments. • Building a dynamic model-routing architecture to cascade inference between fast-response models and deep-reasoning engines to optimize latency and API compute costs. •
  • M
    Melody Generation using RNN(LSTM)
    Jan 2024 - Sep 2024 (9 months)
    • Developed a melody generation model using LSTM for music sequence prediction and utilized Music21 to handle MIDI file generation. • Integrated NVIDIA CUDA Toolkit for faster training on a GPU, reducing training time by 40%. • Built and deployed an interactive web interface with Streamlit on Hugging Face Spaces, allowing users to input melodies, generate sequences, and download them as MIDI files. • Focused on creating a user-friendly music creation experience with real-time feedback.
  • C
    Character Information Extraction System
    Jan 2023 - Aug 2023 (8 months)
    • Designed a system to process text documents into structured formats, compute embeddings with MistralAI, and store them in a persistent vector database (Chroma) for efficient similarity search. • Integrated an LLM to extract detailed character information, generating structured JSON summaries with relationships and roles. • Delivered a scalable pipeline with a user-friendly CLI, enhancing efficiency for embedding computation and character information retrieval. • While the project was developed using Python, LangChain, and Chroma, I can attest to the foundational principles of CI/CD that are implicitly applied in ensuring robust and repeatable deployment processes, even within a Python-centric development environment. • Separated the use o
  • H
    Hostflow
    May 2022 - Jan 2026 (3 years 9 months)
    My technical and professional skills include: Programming: Python, C/C++, Data Structures, Algorithms, SQL, and JavaScript. AI and ML: TensorFlow (Keras), YOLOv8, Neural Networks, NLP, RAG Workflows, LLM fine-tuning, SFT & RLHF pipelines, Sentence Transformers, and LSTM. Tools and Frameworks: LangChain, REST Framework, Docker, CUDA, Next.js 13, React, Prisma, Tailwind, MySQL, Gradio, Pydantic, Typer CLI, and music21. Specializations: Vector Databases (Chroma), Embedding Models (MistralAI), Cloud Computing (AWS, Oracle), Git/GitHub, ARM NEON SIMD, and OpenClaw. Soft Skills: Quick Adaptability, Creative Thinking, Problem-solving, Leadership, and Communication.
  • A
    AI Therapist Chatbot using Sentence Transformers
    Apr 2022 - May 2022 (2 months)
    • Developed an AI therapist chatbot using fine-tuned SentenceTransformer models to provide empathetic, contextually aware responses from a custom dataset. • Created an interactive interface with Gradio, enabling dynamic conversations and features like conversation history tracking and chat clearing for enhanced user experience. • Deployed the chatbot on Hugging Face Spaces, ensuring easy accessibility and seamless interactions. • Applied software design principles to structure the Gradio interface, ensuring a user-friendly and intuitive interaction flow. • Leveraged cloud architecture concepts through the deployment of the AI model to Hugging Face Spaces, demonstrating an understanding of scalable and accessible deployment strategies.
  • P
    B.Tech in AI&DS
    Posted by me
    Apr 2021 - May 2026 (5 years 2 months)
    Programming Python · C/C++ · SQL · TypeScript · Bash AI / ML LLMs · RAG · Agentic systems · LangChain · LangGraph · LlamaIndex PyTorch · Hugging Face Transformers · SentenceTransformers · CLIP (CNN) YOLOv8 · Fine-tuning (LoRA / QLoRA) · Quantization (GGUF) · Prompt Engineering Eval & benchmarking (Terminal Bench v2.0, OS Bench) Backend / APIs FastAPI · Pydantic · Server-Sent Events · REST · LangChain MLOps / Infra Docker · docker-compose · MLflow · Airflow · GitHub Actions CI/CD · pytest · ruff Cloud GCP (Cloud Run, Artifact Registry, Secret Manager, Workload Identity Federation) Hugging Face Spaces · AWS (familiar) Vector / D
  • Personal
    Hostflow
    Personal
    Jan 2021 - Current (5 years 7 months)
    My technical and professional skills include: Programming: Python, C/C++, Data Structures, Algorithms, SQL, and JavaScript. AI and ML: TensorFlow (Keras), YOLOv8, Neural Networks, NLP, RAG Workflows, LLM fine-tuning, SFT & RLHF pipelines, Sentence Transformers, and LSTM. Tools and Frameworks: LangChain, REST Framework, Docker, CUDA, Next.js 13, React, Prisma, Tailwind, MySQL, Gradio, Pydantic, Typer CLI, and music21. Specializations: Vector Databases (Chroma), Embedding Models (MistralAI), Cloud Computing (AWS, Oracle), Git/GitHub, ARM NEON SIMD, and OpenClaw. Soft Skills: Quick Adaptability, Creative Thinking, Problem-solving, Leadership, and Communication. In terms of my core competencies: • Demonstrated proficiency in Python and C/C++