- Shipped an AI call-analytics pipeline from prototype to production in two weeks. It transcribes sales calls and pulls out outcomes, customer intent and commitments automatically, backed by 80 automated tests.
- Increased concurrent call processing 6–10× on the same hardware by redesigning how audio is segmented and processed in parallel.
- Reduced Norwegian transcription errors by about 20% while making transcription twice as fast, by benchmarking speech models on a multilingual test set.
- Kept analysis to about 3 seconds per call by evaluating 8 language models on accuracy, speed and cost.
- Architected a multimodal pipeline that turns long mathematics lecture videos into structured LaTeX documents and PDFs, combining speech transcription, OCR and vision-language models.
- Restored a production service that had stopped loading sessions, recovering 183 GB of disk from more than 2,000 orphaned containers, and added nightly cleanup to prevent repeats.
- Found and fixed analytics maths that overstated customer conversion by up to 44%.
- Cut a product demo's response to clicks from 6.4 seconds to under 50 ms.
- Hardened containerised Linux environments with SSH key authentication and TOTP multi-factor auth, and made server builds reproducible with Packer and Ansible.
- Fixed reliability, security and integration issues across C++ streaming services, Django APIs and JavaScript front ends, including a Google Meet integration.
Saurav Singh Chandel
Software developer focused on AI engineering
Intermediate Software Developer at Studix. Based in St. John's, Newfoundland.
I build applied AI features end to end: speech transcription, LLM analysis pipelines, document understanding with OCR and vision models, and running open-source models on our own hardware.
I studied Computer Science and Applied Mathematics at Memorial University, so I'm as comfortable with the maths behind a model as with the systems that serve it.
Work
- Set up CI/CD for the app.
- Worked on the AI knowledge graphs behind the workspace, and on prompt handling.
- Built the pipeline that turns transcriptions into structured artifacts.
- Ran security testing and hardened the app.
- Owned the database: Postgres on Supabase.
- Built the admin dashboard and metrics.
React, JavaScript, Postgres, Supabase
Teaching Assistant, Mathematics
Sep 2023 – Dec 2024Led tutorials and graded coursework for more than 100 students.
Computer Science Help Centre Assistant
Sep – Dec 2024Helped students debug and understand their programming assignments.
Database Analyst
May – Aug 2022Built Selenium automation to find and fix email delivery failures, saving about 30 hours of manual admin work each semester.
Projects
Education
B.Sc. Joint Honours, Computer Science and Applied Mathematics
Memorial University of Newfoundland
GPA 3.76, and 3.9 in Computer Science. Dean's List 2022–23.
My honours thesis studied when a black hole's boundary is stable, using symbolic and numerical spectral methods in Python and SymPy. The computed eigenvalues matched known solutions to relative errors as low as 10⁻⁵.
Skills
- AI engineering
- LLM integration, speech transcription, OCR and vision models, classification, context management
- Local models
- Open-source model evaluation, local inference, quantization, hardware sizing
- Machine learning
- scikit-learn, Stable-Baselines3, Gymnasium, NumPy, Pandas
- Languages
- Python, C++, TypeScript, JavaScript, Java, SQL
- Systems
- Django, Node.js, React, PostgreSQL, Redis, Docker, Linux, CI/CD
- Mathematics
- Numerical and spectral methods, SymPy, scientific computing
Contact
The quickest way to reach me is email. I'm also on LinkedIn, X and LeetCode.
sauravsingh527@gmail.com