01
SemiRestore
BuiltAn AI-assisted system for restoring degraded, single-channel semiconductor scanning electron microscope images.
Portfolio / 2026Hyderabad, India
Akhilesh Kancharla
Machine Learning · Applied AI · Systems
I work across machine learning, computer vision, data pipelines, and the engineering systems that make experimental work reliable and useful.
Current focus
Reliable vision systems · data quality · navigation under signal loss
01
All projects →Selected work
02
Vantage
In progressA placement-management system that turns irregular academic records into auditable eligibility and recruitment workflows.
03
An applied navigation project exploring smartphone sensor data when GNSS positioning becomes unavailable.
02
Now
Intelligent Dead Reckoning
Useful dead reckoning starts with disciplined sensor acquisition and validation before model complexity is introduced.
03
Research notes →Research
Deterministic pipelines vs. large language models
An experimental comparison of deterministic parsing and LLM-based extraction across clean and noisy inputs, with emphasis on extraction quality, latency, cost, and failure behavior. The work grew from the Vantage student-record ingestion pipeline.
04
Education
Mahatma Gandhi Institute of Technology
B.Tech, Computer Science and Engineering
Undergraduate study in computer science with project work spanning machine learning, backend systems, and applied engineering.
IIT Madras
B.Sc., Data Science
Parallel study in data science, statistics, programming, and computational problem solving.
A useful next conversation
Working on a difficult ML or data problem?
Open to ML internships, research collaborations, and technically ambitious project work.
Start a conversation