Regalia Pass On
A student marketplace for passing academic regalia on to current students. Deployment is coming soon.
Customer problem
Graduation attire is expensive, used briefly, and often left unused while the next class needs it.
Currently building at NewsGenie
Software Engineer - AI
I work backward from real customer problems to build reliable AI products, backend services, and cloud systems that create measurable value.

Measured outcomes
0%
faster analysis
40hr to 2hr per research cycle
0
Hours automated yearly
Multi-LLM review pipeline
0M
records consolidated
Enterprise MDM migration
0%
cloud cost cut
Optimized Lambda container builds
0.0%
uptime held
CloudWatch-backed observability
0
Entities tracked per run
0.82 mAP multi-object tracking
Selected work
I start with what people need, then build the smallest reliable system that solves it.
Building now
Two products I am taking from real user problems to public launch.
A student marketplace for passing academic regalia on to current students. Deployment is coming soon.
Customer problem
Graduation attire is expensive, used briefly, and often left unused while the next class needs it.
A productivity app that uses psychology-informed, real-time nudges to help people recover from distraction. I am continuously improving it for a public launch.
Customer problem
Most productivity tools block distractions without helping people understand and recover from losing focus.
Case studies
Working demos with the architecture, decisions, and tradeoffs behind each system.
Plus 2 more covering financial-systems middleware at CITI Bank and a ten-million-record master data pipeline at Avery Dennison. Browse all work.
Experience
Feb 2026 to Present
currentMesa, AZ
NewsGenie, Inc.
Building a real-time content protection service that lets publishers encrypt digital content and revoke access from bots and scrapers in under a second.
Aug 2024 to Jul 2025
Tempe, AZ
Arizona State University
Replaced a manual geotechnical research workflow with a serverless deep learning pipeline and a multi-object tracker that follows thousands of soil particles across video.
Jan 2023 to Nov 2023
Remote
LTIMindtree · Client: CITI Bank
Automated configuration management for financial systems behind hardened middleware, with SSO, role-based access control, and a full audit trail.
Aug 2021 to Dec 2022
Remote
LTIMindtree · Client: Avery Dennison
Consolidated ten million customer records scattered across regional systems into a single trusted master data set feeding downstream CRM and analytics.
Technical depth
Arizona State University · Tempe, AZ
Dec 2025
Statistical Machine Learning · Cloud Computing · Data Processing at Scale · Software Security · Semantic Web Mining
Nagpur University · India
Jul 2021
Data Structures & Algorithms · Operating Systems · Database Management Systems · Distributed Systems · AI/ML
Recognition
International Collegiate Programming Contest, top 2.4% worldwide.
Recognized for technical excellence and cross-team collaboration.
Statistical Machine Learning, Cloud Computing, Data Processing at Scale.
SoDA, HackerDevils, and GDSC at ASU; active in open source on GitHub.
About
I am a Software Engineer - AI at NewsGenie, where I build a real-time content protection service for publishers. Outside work, I am building Regalia Pass On to help students reuse graduation attire and rebuilding ReFocus.AI, my hackathon productivity prototype, for a public launch.
Previously, at Arizona State University, I replaced a forty-hour research workflow with a serverless deep learning pipeline. At LTIMindtree, I built financial middleware for CITI Bank and worked on a ten-million-record master data migration for Avery Dennison.
I earned my M.S. in Computer Science from Arizona State University in December 2025. My team also placed 239th out of 10,000 teams in ICPC 2019. I am based in Mesa, Arizona and open to relocation.
How I work
I work backward from who needs the product, what is failing today, and what a useful outcome looks like.
Bounded retries, re-identification after occlusion, and refusing to decrypt before classifying all address what happens when something goes wrong.
RAGAS evaluation on the agent, validation inside the ETL pipeline, SonarQube in the merge path. Standards that run automatically are the ones that hold.
Docker and Terraform removed config drift from a research pipeline whose conclusions depended on being re-runnable.
Recommendation
I worked with Akash for 2 years at LTIMindtree, and he stood out from day one. He ramped up quickly on a complex Customer Data Management project, mastered Oracle SQL and Fusion CDM architecture, and delivered solutions that reduced client operational costs by 20%, which led to follow-on business.
Akash is not just technically sharp, he takes full ownership, asks thoughtful questions, and mentors teammates effectively. He is reliable, adaptable, and thrives under pressure. I recommend Akash without hesitation. He is the kind of engineer who raises the bar.
Sarthak Mohanty
AI Evangelist @ ITM | Agentic AI, Generative AI, AI Data Platforms & Analytics | Delivering AI-infused solutions with a strong focus on Responsible AI
Managed Akash directly at LTIMindtree | Aug 31, 2025
Contact
Happy to walk through any of the architecture decisions in detail, including the ones that did not work out. Fastest way to reach me is email.