Rochester Institute of Technology
Aug 2024 to Dec 2026Master of Science, Computer Science · Rochester, NY
Prakyat Prakash
Software engineer and machine learning researcher
Summary
Software engineer and pipelines architect by practice: data pipelines, full-stack systems, and ML models that go from notebook to production. I also do research, on proteomics and how the genome regulates itself.
Education
Experience
Graduate Research Assistant, Machine Learning
May 2025 to presentRochester Institute of Technology · Rochester, NY
- Engineered an end-to-end Python ETL pipeline to extract, clean, transform and validate 24.4M raw peptide-spectrum matches from large biological databases, producing 361,789 experimentally verified phosphosites across 35 parallel workers.
- Developed scalable data transformation workflows and structured data models to integrate heterogeneous biological datasets, with data quality checks, deduplication and FASTA verification for consistency and reliability.
- Designed data engineering solutions for large-scale biological data processing, optimizing parallel execution, database integration and reproducible pipelines for downstream machine learning.
AI Engineer, Contract
Jan 2025 to Apr 2025Acinonyx Technologies Pvt. Ltd. · Remote, NY
- Developed Python data processing and transformation pipelines for LLM-powered RAG applications, improving code generation accuracy by 25% through retrieval optimization and prompt engineering.
- Implemented structured data retrieval, database integration and automated evaluation workflows using LangChain, reducing model output errors by 30%.
- Applied SDLC principles, Agile methodologies, Git-based version control and CI/CD practices to deliver scalable AI data solutions.
Research
Getting signal out of noisy biology
I have worked in two labs at RIT, the KC Lab and the Cui Lab, and I am a Graduate Research Assistant in the Cui Lab. The work is machine learning and deep learning on life science and genomics: which computational methods survive contact with a real experiment, and which quietly learn the shape of the experiment instead.
Two threads right now. Building rigorous extraction pipelines for post-translational modification data, which meant correcting an assumption the field had accepted for years. And studying genomics and 3D chromatin organization with deep learning models.
Publications
Machine learning-based determination of sex-related bladder cancer biomarkers
2026Intrinsic DNA codes govern distinct modes of nucleosome and transcription factor interactions
2026Projects
Live dashboard ↗
Scrapes every World Cup match from ESPN, finds the exact hydration break timestamps, and measures attacking pressure either side of the whistle. Pressing teams lose momentum across the break; the teams being pressed gain it. Rebuilds itself daily.
Live dashboard ↗
Rebuilt the training data for phosphorylation site prediction. The field calls every unreported residue a negative; this verifies them instead, turning 24.4M peptide-spectrum matches into 361,789 checked sites and cutting a trypsin sampling bias from 33.2% to 3.6%.
Skills
Languages
Python · SQL (PostgreSQL) · Java · C++ · JavaScript
Data engineering
ETL pipelines · Data transformation · Data integration · Data warehousing · Data lakes · Apache Spark
Cloud and databases
AWS (S3, DynamoDB, Lambda, API Gateway) · Neo4j · PostgreSQL
Software engineering
SDLC · Agile · Docker · Git · CI/CD · REST APIs
Machine learning
PyTorch · TensorFlow · Scikit-learn · Pandas · NumPy · LangChain · RAG
Contact
Open to 2026 SWE and ML roles. Based in Rochester, NY.