/ résumé
Download PDF
Sameer Surla
Data Engineer @ FedEx · GenAI & Agentic AI
sameersurla07@gmail.com +91-7219736248 github.com/sameersurla-iitm linkedin.com/in/sameer-surla-iitm

Experience

Data Engineer I — FedEx, Bengaluru
Jul 2025 — Present
US Surcharge Dashboard — SAS → Ab Initio Migration · US RQ Team
  • Led design and development of a production-grade ETL pipeline processing 113M+ rows (25 months of data).
  • Built a config-driven solution using Ab Initio PDL, reducing rate-update turnaround from 10 days to 4 hours.
  • Ensured data accuracy and consistency through large-scale validation of historical datasets.
  • Developed an automated reporting workflow for backend data extraction and email distribution.
Artificial Intelligence Intern — Detect Technologies, Chennai
Jan 2025 — May 2025
  • Integrated Gen-AI for industrial safety via end-to-end image generation and real-time video analysis pipelines.
  • Designed a customizable image generator with Hugging Face's FLUX model (GUI and non-GUI modes).
  • Developed a logging framework to capture detailed prompt and configuration data for traceability.
  • Deployed multi-threaded video analysis using Qwen VLM, OpenCV & Streamlit for real-time safety monitoring.
Artificial Intelligence Intern — Diax.AI, Chennai
May 2024 — Jul 2024
  • Developed a deep-learning model to detect diabetic foot ulcers, now adopted by the Tamil Nadu State Government.
  • Achieved 95% accuracy in YOLO-based ulcer detection, with potential impact over 10M patients.
  • Enhanced model robustness by 15% through annotation and augmentation of 3,000+ images using Roboflow.
  • Led the full software lifecycle: preprocessing, training and deployment.
SDE & Reliability Intern — MachIntell, Chennai
Feb 2024 — May 2024
  • Developed solutions for OEMs to detect and resolve machine malfunctions for reliability engineering.
  • Achieved 98.76% accuracy on a hardware product for sensor and actuator monitoring.
  • Reduced diagnosis time by 40% via real-time malfunction analysis using MCF, Laplace and event plots.
  • Integrated detection across database, frontend, backend and hardware teams.

Projects

Atlas — Autonomous Research Agent  [case study] [live demo]
  • 4-agent autonomous research pipeline (Planner→Researcher→Critic→Synthesizer) doing live web research, self-critique loops, and streaming fully-cited reports via Gemini & Groq LLMs.
  • Cross-model failover with rate-limit pacing across free LLM tiers — $0 cost, 8+ cited sources and verified telemetry per run; source-grounded extraction eliminates hallucinated citations.
RAG-Powered Multi-Doc Q&A Chatbot
  • FAISS indexing + HuggingFace embeddings; integrated Cohere LLM for 95% multi-doc accuracy; RCTS chunking cut latency 35%.
Lunar Lander — Reinforcement Learning
  • DQN agent in OpenAI Gym with experience replay + target networks; reached avg reward ~250, surpassing the solved threshold (200) within 800 episodes.
Credit Card Default Prediction
  • XGBoost at 90% validation accuracy over 300K+ transactions; Gradient Boosting at 86.3% F1 and 0.870 AUC.
Fantasy IPL Prediction Hackathon — AI Club IITM
  • 87% accuracy and 0.87 AUC on 22K+ profiles using TensorFlow; reduced WMAE by 90% with RF/XGBoost and stacking.
Deep Learning Transliteration Model — RNN
  • LSTM model for Hindi→English transliteration on 200K+ sentences; improved accuracy 76.8% → 90.6%.

Education

B.Tech, Civil Engineering — IIT Madras
2021 — 2025
CGPA 7.92 / 10.0
Class XII (PCM) — Narayana Junior College, Mumbai
2021
95.67%

Skills & Certifications

Languages: Python, SQL, C++, MATLAB, HTML, CSS
AI / ML: PyTorch, TensorFlow, Keras, Scikit-learn, XGBoost, OpenCV, NLTK, OpenAI Gym
GenAI / LLM: RAG, Agentic & Multi-Agent Systems, Prompt Engineering, FAISS, LangChain, Hugging Face, Cohere, VLMs, Embeddings
Data Engineering: Ab Initio, PDL, SAS, ETL, Teradata, Data Validation
Tools: Git/GitHub, Pydantic, Streamlit, Jupyter, Google Colab, Roboflow, Tableau, LaTeX, Kaggle
Certifications: Deep Learning, ML & Algorithms (Stanford); Data Analytics (Google)