%%%%%##                    SAHIL CHAUDHARY
              #%@@@@%%%%%                  -------------
              %###**++*#%                  OS: ..................... Windows 11 / Linux
              %*++==+++*%                  Education: ............... B.Tech AI & Analytics, GLA University
              *+#%*+##*=**                 Visiting: ................. IIT Gandhinagar (CSE)
              +=+++=++==+*                 CPI: ...................... 9.06/10.0 | 9.14/10.0 (IITGN)
              =+*#*#*#**                   
                *#*****=                   Lang.Programming: ......... Python, Java, JS/TS, SQL
                =#####*..                  Lang.ML: .................. TensorFlow, PyTorch, Scikit-learn, HF
             :. .*####+ .:..               Lang.Web: ................. React, Node.js, Express, MongoDB, I love System Design
         %-.:... -%##%. -....-%%#          
      %%@*..:...  #*#+ .:.... =@%%         Focus: .................... ML & Software development, XAI, Full Stack
     %%@@-.:::... -++.  :...:.:@@@%        Competitive Programming: .. 700+ solved, CodeChef Rank 169/20k+
    %@@@#:.:::.:. -+=.  :...:::%@@@@       
    @@@@*:::::.::.+*+-..::::-::%@@@@       Contact ------------------------
   @@@@@+:-:-:::=-+*+=-:-:----:#@@@@@      LinkedIn: ................. Sahil--Chaudhary
  @@@@@@*-------==+*#+=:----=-:%@@@@@@     Email: .................... sahil.chaudhary7721@gmail.com
  @@@@@@%-------++*#%*+=+**+=--@@@@@@@     Portfolio: ................. dhritam.com
 @@@@@@@%=--==++**###****%%*===%@@@@@@@    
 @@@@@@%*+++***##*++==+*%@#****++%@@@@@@   GitHub Stats ------------------------
@@@@@@@***#####**-::...:+**######%@@@@@@   Repos: .................... (live badge below)
@@@@@@@%%%%##*+==---.   :--=+**@@@@@@@@    Streak: ................... (live badge below)
 @@@@@@@@#+=-=====--.   .------*@@@@@@     Top Langs: ................ (live badge below)
        +=--::::-:...   .-:::-:-           

Building things that need to actually work — and then figuring out why they work.

Building end-to-end ML/NLP pipelines (entity extraction, sentiment analysis, feature engineering) on large-scale business text for empirical marketing analytics research. Shipping production-grade features across a React frontend and Node.js/MongoDB backend for a live, user-facing product.

UrjaLink — ML-Powered Microgrid Energy Optimization · Python TensorFlow MongoDB IoT · SIH 2025 Finalist Led backend for a 6-person team; LSTM/ARIMA demand forecasting at 92% accuracy and Dijkstra's/A* power routing across 5+ IoT nodes, cutting simulated downtime 40%. Node.js/Express + MongoDB backend serving 500+ concurrent users with a 35% throughput improvement.

XAI for Transformers — Conservative Attention Attribution · PyTorch BERT GPT · Independent Research Proposed a continuous γ ∈ [0, 1] parameter that adaptively interpolates between gradient-based and conservation-preserving attention attribution, letting practitioners trade off faithfulness for interpretability instead of picking one. Validated via ablation studies on BERT and GPT.

Brain-Computer Interface Pipeline · Python NumPy SciPy SVM Neural Networks · HAIx Lab, IIT Gandhinagar Built a full BCI pipeline combining neural signal processing (FFT, bandpass filtering, feature extraction) with Tobii eye-tracking hardware, reaching 85%+ decoding accuracy across 50+ real experimental trials and cutting inference latency by 40% for real-time decision communication.

Grant Application Automation Pipeline · Python LinkedIn API OpenAI API · Invesho AI, Berkeley Engineered an automation pipeline processing 500+ profiles, validated through A/B testing, that improved grant submission throughput by 300% (2 to 7–8 submissions/day) and lifted user acquisition by 20%.

Telegram Bot for SSB Aspirants · Python Gemini API Redis Deployed to 3,500+ active users with a 40% engagement rate; Redis caching cut response latency 65% and API cost 50% versus GPT-4.

First Place, First Global AI Summit, Noida: ₹50,000 prize for a production-grade AI solution built under competitive conditions. Selected, PayPal Career Academy 2.0 (Tech Track): 1 of 47 candidates chosen pan-India for this fintech mentorship program. Competitive Programming: 600+ problems solved (LeetCode, Codeforces, CodeChef); Global Rank 169/20,000+ (CodeChef Starters 211); Institute Rank 1 (Spectra 3.0). 14x Hackathon Winner/Finalist, including Top 5 (Azure Microsoft, 500+ teams, All India), Top 40 (Hack4Health by Horiba), and SIH 2025 Finalist. Leadership: Tech Lead, GDG Core Team (200+ students); Data Lead, MLSA (6,000+ applicant hackathon); Engagement Lead, ML Core Team (85% retention).