Skip to content
View tanaybaswa's full-sized avatar
⚑
Always Improving
⚑
Always Improving
  • Enkrypt AI
  • Berkeley, California
  • 13:13 (UTC -05:00)

Block or report tanaybaswa

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
tanaybaswa/README.md

Hi there! πŸ‘‹ I'm Tanay Baswa

πŸš€ About Me

  • πŸ” Founding AI Researcher and Engineer specializing in AI Safety and Security.
  • 🌱 Currently working on Multimodal AI and Agentic Evaluations.
  • ⚑ Previously focused on guardrails, safety alignment, and AI reliability systems for high-fidelity industries.
  • πŸ›  Passionate about ensuring AI systems are safe, secure, and robust.

πŸ”₯ Other Work

1. LLM Safety Leaderboard

  • Built an interactive leaderboard to evaluate and compare LLM safety metrics.
  • Features real-time performance and risk-based evaluation.

2. Policy-Based Guardrails and Violation Detection

  • Engineered a system for detecting and handling policy violations in AI systems.
  • Enabled robust decision-making through clear, automated rule enforcement.

3. Confidential Computing Container-Based Solution

  • Designed secure containers to protect sensitive AI data and models during execution.
  • Leveraged cutting-edge confidential computing techniques for end-to-end encryption.

4. Semantic Cache Solution

  • Developed a caching mechanism that uses semantic similarity for faster and more efficient query resolution.
  • Improved latency and reduced computational overhead for large-scale systems.

5. AI Anti Hallucination System

  • Created an automated detection and mitigation system for AI hallucinations.
  • Advanced the reliability of AI outputs in critical use cases.

πŸ› οΈ Skills

Programming Languages

  • Python, TypeScript, JavaScript, C, C++, C#, Java, Go, Rust, Bash, R, SQL, Scheme, Lua (Kong)

Tools

  • Docker, Kubernetes, Terraform, Git, Supabase, Firebase, AWS (SageMaker, EC2, S3, Lambda), Azure (VMs, Blob Storage), GCP (BigQuery, Vertex AI), RunPod, Vercel, Shadeform

Frameworks & Libraries

  • AI/ML: PyTorch, TensorFlow, Hugging Face Transformers, scikit-learn, XGBoost, LightGBM, OpenCV, fastai, vllm, trl
  • Web Development: React, Next.js, Vue.js, Flask, FastAPI, Django,
  • Frontend Styling: Tailwind CSS, Bootstrap, Material-UI, Streamlit
  • Data Engineering: Apache Spark, Apache Kafka
  • Databases: PostgreSQL, SQLite, MongoDB, Redis, DynamoDB, Elasticsearch, ChromaDB, Milvus, Qdrant
  • DevOps: Prometheus, Elastic Stack (ELK), Helm
  • Testing: Pytest, Selenium, JUnit, Postman

🌐 Connect With Me

LinkedIn Twitter Portfolio


πŸ† Achievements

  • πŸš€ Built cutting-edge safety alignment systems for AI reliability.
  • 🌟 Produced guardrails for high-stakes industries.
  • πŸ“– Exploring the frontiers of AI safety research.

Pinned Loading

  1. Database-System Database-System Public

    An implementation of an efficient and lightweight database system that utilises B+ tree indices, efficient join algorithms, query optimization, multigranularity locking to support concurrent execu…

    Java 2

  2. Lightweight-Personal-Operating-System Lightweight-Personal-Operating-System Public

    C 1

  3. OS-Tools-and-Features OS-Tools-and-Features Public

    1

  4. Pet-Adoption-Finder-UI Pet-Adoption-Finder-UI Public

    A front end for a pet adoption finder service that can be connect to a back end for full functinality.

    JavaScript 1

  5. Spotify-Client Spotify-Client Public

    JavaScript 1

  6. MercuryCrypto MercuryCrypto Public

    JavaScript 1