Aaditya A.

Co-Founder

Bengaluru, Karnataka, India6 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in AI and Machine Learning with impactful solutions.
Proven track record in building scalable systems.
Strong foundation in software development and distributed systems.

Stackforce AI infers this person is a SaaS-focused AI Engineer with expertise in Machine Learning and Computer Vision.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Machine LearningComputer VisionSoftware Development

Other Skills

FastAPIPythonSDVGTCFakerFHIROntology EnginesNeo4jChromaDBQdrantMongo AtlasLangChainLlamaIndexOpenAIGoogle ADK

About

I am a passionate Machine Learning and Deep Learning engineer with a strong foundation in AI, distributed systems, and software development. Skilled in Python, Java, C++, and Go, I have experience working with frameworks like TensorFlow, PyTorch, and Flask, along with tools such as Docker, Git, and MLflow. My expertise spans model development, deployment, and optimization, with a keen interest in leveraging AI for impactful solutions. Actively involved in hackathons and research, I enjoy solving complex problems and building scalable, efficient systems. Always eager to learn, I thrive in dynamic environments that push the boundaries of innovation.

Experience

6 mos

Total Experience

6 mos

Average Tenure

Current Experience

Kronosx ai

Founding AI Engineer (Agentic AI)

Sep 2025 – Mar 2026 · 6 mos · Bengaluru · Hybrid

Tech Stack:-
FastAPI, Python, SDV (CTGAN/TVAE/Diffusion), GTC, Faker, FHIR/Da Vinci PDex, Ontology Engines, Neo4j, ChromaDB, Qdrant, Mongo Atlas, LangChain, LlamaIndex, OpenAI/Groq, Google ADK, LiveKit, LiteLLM, Docker, Nginx, AWS EC2/S3, Terraform
Focus Areas:-
Ontology-driven synthetic data, privacy-preserving healthcare data generation, multi-agent automation, FHIR interoperability, statistical/structural fidelity scoring, metadata intelligence, cloud-native AI deployments.
Key Highlights:-
Built an ontology-aware anonymization + synthetic data pipeline achieving 92–97% statistical fidelity and 100% referential consistency.
Automated schema + semantic type detection, cutting manual cleaning by ~65%.
Designed a multi-agent planner/worker/validator workflow with deterministic routing, improving reasoning accuracy by ~30%.
Integrated Neo4j + ontology mappings for cross-entity retrieval speedups (250ms → ~40ms).

FastAPIPythonSDVGTCFakerFHIR+17

Macv ai

Computer Vision Intern

May 2025 – Sep 2025 · 4 mos · Bengaluru, Karnataka, India · Remote

Tech Stack: FastAPI, Streamlit, MongoDB, OpenAI Whisper, GPT-4, LLaMA 3, OpenAI TTS, Agora WebRTC
Focus Areas: Intelligent voice conversations, real-time alert processing, autonomous agent interaction
Key Responsibilities:
Designed and developed full-stack architecture using FastAPI, Streamlit, and MongoDB
Integrated speech-to-text (OpenAI Whisper), LLMs (GPT-4, LLaMA 3), and text-to-speech (OpenAI TTS) for live AI-driven voice communication
Built a structured agent communication pipeline supporting voice, text, and media responses linked to alert IDs
Implemented real-time voice communication using Agora WebRTC for browser-based calls
Optimized deployment for edge devices (e.g., Jetson Xavier NX) with offline inference and low-latency performance

FastAPIStreamlitMongoDBOpenAI WhisperGPT-4LLaMA 3+4

Brandnav

Python Developer Intern

Jan 2025 – Mar 2025 · 2 mos · Remote

Hinglish Speech-to-Text System
Developed an innovative transcription system leveraging open source models (Whisper, Vosk, and others) for mixed Hindi-English audio with Indic Transliteration.
Built a lightweight solution that runs efficiently on low-edge devices like Raspberry Pi, ensuring high transcription quality for both real-time and batch processing through a scalable FastAPI backend with WebSocket streaming.
Face Detection & Keyframe Extraction
Engineered a robust video analysis pipeline leveraging InceptionResnetV1 and Insight_Face for real-time face detection.
Implemented a histogram-based keyframe extraction method, with processing times ranging from ~30 to ~150 seconds (InceptionResnetV1) and ~60 to ~220 seconds (Insight_Face), while consistently extracting about 38–39 keyframes.
Image Search Platform & API
Designed a RESTful API with Flask to manage image embeddings using OpenAI’s CLIP and PostgreSQL (pg_vector) for rapid similarity searches.
Developed a user-friendly Streamlit interface, ensuring a seamless, efficient end-to-end image search experience.