Associate AI Engineer at Centific | AI/ML • Backend • Cloud | Hyderabad, India
Achievement-driven Associate AI Engineer with 1.2+ years of production experience building scalable, cloud-native backend services at Centific. I design, deploy, and operate high-availability AI and data services used in enterprise environments — with hands-on experience across LLM fine-tuning, RAG pipelines, agentic AI, and distributed systems.
- 🏆 Star Award (Q3) at Centific for outstanding contributions to LLM fine-tuning and AI platform development
- 🎓 B.Tech in AI & Data Science, CGPA: 9.12 — Chaitanya Bharathi Institute of Technology
- 📍 Shamshabad, Hyderabad, Telangana
| Category | Technologies |
|---|---|
| Languages & Frameworks | Python, Java, C, Flask, FastAPI, SQL |
| AI/ML | LLM Fine-tuning (SFT, DPO, RLHF/PPO), Hugging Face Transformers, PyTorch, MLFlow, OpenCV, CLIP |
| Agentic AI | LangGraph, RAG Pipelines, Multi-Agent Systems |
| Vector Databases | Pinecone, FAISS, Databricks Vector Search, Milvus |
| Cloud & DevOps | Azure (Blob, AKS, Key Vault), Docker, Kubernetes, Apache Ray, RunPod, Denvr |
| Databases & Logging | PostgreSQL, MongoDB, ELK Stack (Elasticsearch, Logstash, Kibana) |
| Tools | Git, Jira, SonarQube, Postman, DBeaver, VS Code |
- Built Agentic AI solutions using LangGraph for scalable single-workflow automation across multiple AI use cases
- Architected and deployed 15+ LLM fine-tuning AI services (Qwen-32B, QwQ-32B, Microsoft-Phi4, NVIDIA VILA, YOLO v11) with SFT, DPO, and RLHF/PPO support
- Engineered MLFlow integration with 20+ evaluation metrics (BLEU, ROUGE, F1, toxicity, perplexity) for model benchmarking
- Led compute offloading implementation for on-premise edge deployments with GPU optimization
- Developed Apache Ray POC for distributed parallel execution of AI services
- Implemented dataset validation and splitting logic across 8 LLM models
- Built RAG pipelines using Pinecone and Databricks vector databases for multi-modal data processing
- Developed video chunking and embedding system using CLIP + OpenCV for similarity search across text, image, and video
- Implemented ELK logging for 10+ AI services enabling centralized monitoring and debugging
- Designed RESTful APIs using FastAPI/Flask integrated with PostgreSQL for metadata storage
AI Workbench - LLM Fine-tuning Platform | Centific
End-to-end fine-tuning services for enterprise LLMs including Qwen-32B, QwQ-32B-Preview, and Microsoft-Phi4 — supporting SFT, DPO, and PPO methodologies with custom tokenizer integration.
Python Hugging Face MLFlow Azure Blob RunPod Docker Kubernetes
RAG Studio - Multi-Modal Pipeline | Centific
Vectorization and indexing pipelines for RAG applications supporting text, image, and video data with CLIP embeddings.
Python CLIP OpenCV Pinecone Databricks FAISS Flask PostgreSQL
Frontier AI Radar | Hackathon Project
Multi-agent AI pipeline that monitors AI research papers, leaderboards, and benchmarks — delivering automated PDF digests via email.
Python FastAPI PostgreSQL Claude API SendGrid ReportLab
Open-source event planning tool with email communication, secure login (AUTH0), and automated attendance tracking — achieved 40% increase in event registration efficiency.
React JS Node JS Express JS AUTH0 MongoDB
Offline-capable Android app using OCR (MLKit) for text extraction from images with audio feedback via TTS.
Android Studio MLKit Java OpenCV CameraX TTS
- 🌟 Star Award (Q3) at Centific — for exceptional contributions to LLM fine-tuning services and AI platform development
- 🥇 Centific Premier Hackathon winner — two-week Data, AI & ML hackathon developing a workflow optimization platform
- 📜 Stanford University (Coursera) — Unsupervised Learning, Recommenders, and Reinforcement Learning
- 🚀 ISRO — Overview of Space Science and Technology training program
- 📖 India Book of Records holder — Smallest Replica of Telangana Amaraveerula Stupam on Pencil Lead
- 💻 Solved 300+ DSA problems on LeetCode and competitive programming platforms
"Building AI systems that scale — one fine-tuned model at a time."