Open to opportunities

Akritah Sahu

@akritahsahu

Message

ML-focused software engineer shipping RAG and LLM evaluation for hallucination risk and reliable retrieval.

India

Message

What I'm looking for

I’m looking to build reliable LLM/RAG systems with measurable quality—real-time hallucination risk monitoring, retrieval consistency, and production-grade evaluation (F1, calibration, latency/cost). I enjoy teams that value auditable outputs and end-to-end ownership.

I’m a 2nd-year CS undergrad who’s shipped end-to-end ML systems with RAG pipelines, LLM evaluation, document ingestion, and production-minded deployments. I’ve built hallucination detection and financial AI products under real competition and internship conditions, with a research focus on real-time hallucination risk in LLMs using retrieval consistency monitoring.

Across my projects, I decompose LLM outputs into atomic facts, retrieve multi-source evidence via hybrid search, and add evaluation harnesses that benchmark F1, calibration, hallucination rate, latency, and cost. I also enjoy building full-stack tools for multilingual NLP workflows and interactive ML demos—always emphasizing auditable, measurable behavior rather than one-off outputs.

Experience

Work history, roles, and key accomplishments

Current

Real-Time LLM Monitoring

Current

Bennett University

Building an inference-time monitoring framework for real-time hallucination risk in LLMs using token distribution uncertainty signals and retrieval consistency tracking for RAG integration.

LLM Evaluation RAG Uncertainty Signals Token Probability Monitoring Retrieval Consistency F1 Benchmarking

Software Engineering Intern

Avijo

Jun 2025 - Aug 2025 (2 months)

Shipped production-ready React components across 3 platforms, improving component reusability and resolving critical rendering bugs in a live codebase.

React TypeScript JavaScript Git REST APIs

Education

Degrees, certifications, and relevant coursework

Bennett University

Bachelor of Technology, Computer Science & Engineering

2024 -

Grade: CGPA: 8.41/10

Activities and societies: IEEE WIE, AI Society, AWS Cloud Club — Core Technical Team

Pursuing a B.Tech in Computer Science & Engineering at Bennett University (2nd year) with a focus on LLM evaluation, RAG systems, and real-time hallucination risk monitoring.