SHUBHAM KESHARI
AI Engineer @ Newron.ai | ex-SDE @ Jio | BITS Pilani | Backend, Cloud and ML Developer | Open to New Opportunities
Bengaluru
3+ years of work experience
Dec 20259 mosPresent
Bengaluru
– Led the end-to-end architecture and delivery of Crediverse, a scalable AI-driven loan processing platform used by
NBFC clients like Aditya Birla Finance and Credila, unifying legacy systems into a single extensible monorepo.
– Architected a configurable Stage Manager Engine to enable dynamic, client-specific workflows by decoupling
business logic into reusable stages and routines, leveraging RabbitMQ, Redis-based rate limiting, and transactional
state management to ensure reliability, retries, and workflow flexibility.
– Handled platform reliability and deployments by Dockerizing services, adding GitHub Actions CI/CD, setting up
Nginx with SSL, and replacing polling APIs with webhooks.
– Set up full-stack observability using SigNoz (OpenTelemetry) for distributed tracing, metrics, and dashboards;
added RabbitMQ DLQs with Slack alerts and trace propagation across services.
Aug 20231 yr 11 mosJun 2025
Navi Mumbai, Maharashtra, India
- Built a session microservice for 10K+ RDP/SSH connections using Apache Guacamole, Spring Boot, and PostgreSQL.
- Implemented secure session logging, idle timeout, and log archival with encryption and retention policies.
- Developed SCP-based file transfer for SSH with recursive navigation and secure upload/download.
-Integrated SAML SSO with Azure AD and TOTP-based 2FA, enabling federated login across services.
- Designed a cross-database optimization engine and Redis config suite supporting 100+ servers.
- Created a no-code Appium script automation tool with performance profiling and test session visualization.
Jul 20226 mosDec 2022
Bengaluru, Karnataka, India
– Developed a WhatsApp notification system for flight and hotel bookings, integrating with email and SMS data
to improve real-time update efficiency and system responsiveness by 50%.
– Enforced templated WhatsApp messages and automated PDF generation, utilizing AWS S3 for storage, and
used Kafka to produce messages to a Logging microservice for seamless integration with third-party services,
collaborating with Yellow.ai as the WhatsApp API provider.
– Debugged and wrote APIs to create functionalities, ensuring reliable operation and diagnosing issues using logs.
May 20213 mosJul 2021
Bangalore Urban, Karnataka, India
– Built a ticket classification system using BERT embeddings and a Multi-Layer Perceptron, achieving
96% accuracy in categorizing 10K+ queries across IT Support, DevOps, Engineering, and QA using stratified cross-validation.
– Trained a MobileNet-based CNN with custom dense layers on 5K+ labeled images from HR, Finance, and IT, enabling 92% accurate classification of unstructured media across cloud repositories.
– Delivered a scalable inference pipeline via FastAPI, integrating model storage on AWS S3 and serverless execution through AWS Lambda, reducing latency to <250ms and supporting stateless, on-demand ML model serving.
SHUBHAM’s AI Stack
Education
Bachelor of Engineering - BE, Electronics and Instrumentation
Birla Institute of Technology and Science, Pilani
20194 yrs2023
Master of Science - MS, Computer Science
Stockholm University
20252025
Join SHUBHAM and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss