Abhilash Bora
Building intelligent systems at the intersection of research and production.
// ABOUT
I'm a Machine Learning Engineer currently building production ML systems at an early-stage startup. My work sits at the intersection of research and engineering taking state of the art models and making them work reliably at scale.
I'm interested in deep learning and building intelligent systems. I like working on large neural networks, training infrastructure, and applying machine learning to real-world problems.
// EXPERIENCE
Machine Learning Engineer
@AiStrikeJan 2025 – Present
- Scaled the company's core AI product 50x in throughput while maintaining full production reliability across all customers
- Owned the end-to-end build and delivery of the AI investigation platform, from infrastructure to customer-facing features
- Drove measurable cost reduction and operational efficiency through observability tools like Langfuse and automation across production environments
Research Intern
@IIT GuwahatiMay 2024 – July 2024
- Researched BERTopic with variational autoencoders for richer latent topic representations under Prof. Dr. Ashok Singh Sairam
- Achieving a 5–10% coherence score improvement over baseline across a corpus of 19k documents.
Research Intern
@NUS School of Computing, SingaporeDec 2022 – Jan 2023
- Designed and trained Bidirectional LSTM models for sentiment analysis on Twitter records
- Led a 5-member team to surpass baseline accuracy by 15%; receiving faculty recognition at NUS for model innovation.
// EDUCATION
B.Tech in Computer Science and Engineering
Indian Institute of Information Technology, Pune
- Relevant Coursework: Data Structures & Algorithms, Machine Learning, Deep Learning, Linear Algebra, Database Systems, Operating Systems, Computer Networks, Software Engineering, Object-Oriented Programming
// PROJECTS
Blockchain-powered platform for tamper-proof medical record management using ERC-721 NFTs on Ethereum; improved data accessibility and auditability for patients and providers, reducing record retrieval time by 30%+
Real-time stock prediction system combining LSTM for price forecasting and FinBERT for financial sentiment analysis; achieved a 40% reduction in end-to-end inference latency enabling faster market response times for traders
Astra is an AI-driven customer support agent that leverages LangGraph to create a 11-stage workflow orchestration and NLP pipelines to deliver fast, personalized, and context-aware assistance.
// SKILLS
Languages
ML/DL
MLOps & Infra
LLM & GenAI
Cloud
// CONTACT
Let's Build Something.
Always exploring new ideas in AI and software development, open to connecting and exchanging thoughts with others in the field.