Senior AI/ML Engineer @UnifyApps | Agentic AI | GenAI | ML | Backend | NSUT’24




Pune District

Germany

New Delhi, Delhi, India







Bachelor of Technology - BTech
Netaji Subhas University of Technology

PCM
City Montessori School
Join Suraj and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal BossSenior AI/ML Engineer @UnifyApps | Agentic AI | GenAI | ML | Backend | NSUT’24
“If intelligence is a cake, the bulk of the cake is unsupervised learning.” — Yann LeCun I care about the bulk of the cake and the infrastructure that bakes it. I work at the intersection of AI and Backend Engineering, turning research into systems that scale. My focus: LLMs, real-time data pipelines, Agentic AI and GenAI architectures built for production. 🔗 Open to connecting, let’s build something.
New Delhi
claude code
hermes agent
hugging face
langgraph
Guardrails AI
vLLM
Senior Platform Engineer at UnifyApps, Sep 2026 - Present · 2 mos
Platform Engineer at UnifyApps, Jul 2026 - Aug 2026 · 2 mos
Product Engineer at UnifyApps, Dec 2025 - Jun 2026 · 7 mos
Software Developer (AI & Backend) at Attentions AI, Aug 2024 - Nov 2025 · 1 yr 4 mos, Pune District, • AI Agents & Automation - Pioneered an autonomous Invoice Processing Agent with OCR, data extraction, normalization, and anomaly detection, reducing manual errors by 85% and improving efficiency. - Built an autonomous Jira Orchestration Agent using agentic workflow automation to manage and execute complex operations, increasing process speed and efficiency by 70%. - Architected a suite of AI-powered autonomous agents using CrewAI and LLM-based frameworks, implementing advanced prompt chaining, context-aware reasoning, and dynamic API orchestration to enable end-to-end business process automation and scalability. • Enterprise Document Search & RAG - Engineered a robust data ingestion pipeline leveraging VisionLM, PyMuPDF, OCR, and LLM-based extraction, integrating outputs into Solr and Milvus for knowledge management. - Improved retrieval relevance by 55% using query expansion, semantic chunking, and hybrid (dense + sparse) search with LLM re-ranking. - Designed an LLM-driven query routing system to differentiate conversational and RAG contexts, enhancing accuracy and reducing response latency by 70%. • Backend & Microservices - Developed a real-time notification service using Redis Streams and MySQL, supporting concurrent clients, retries, and offline delivery for high reliability. - Built a secure JWT-based Auth Service with Google, Outlook, and LDAP integrations via OAuth 2.0/OpenID Connect, featuring refresh tokens, RBAC, and session handling. - Created a scalable File Service using AWS S3 and MySQL for secure uploads/downloads with pre-signed URLs, lifecycle policies, and encryption. - Deployed microservices, Airflow jobs, and APIs within a Dockerized architecture to enhance scalability, maintainability, and deployment efficiency.
Artificial Intelligence Intern at Stealth Startup, May 2024 - Jul 2024 · 3 mos, Germany, • Developed a robust pipeline for real-time audio data processing, which involved parsing data from web sockets, connecting parsed data to LLM, and delivering audio and text responses to the corresponding client. This process significantly improved the efficiency and responsiveness of our client communication system. • Implemented a Short-Term Memory system for chat storage, enhancing the system’s ability to recall and process information. Additionally, utilized Milvus, a powerful open-source vector database, for efficient vector storage, contributing to the overall performance and scalability of the system.
Machine Learning Intern at Be Here Elite Technologies Pvt. Ltd., May 2023 - Jul 2023 · 3 mos, New Delhi, Delhi, India, • Spearheaded the seamless integration of a cutting-edge face detection model into a mobile app, revolutionizing the app unlock process; leveraged deep learning techniques and computer vision algorithms to enable robust and real-time facial recognition, resulting in an 80% reduction in unauthorized access attempts. • Collaborated closely with cross-functional teams of software engineers and UX designers to seamlessly incorporate the face detection feature into the app’s existing authentication system.
Bachelor of Technology - BTech, Netaji Subhas University of Technology
PCM, City Montessori School
github: https://github.com/suraj-singh-sde
suraj-singh-sde.github.io: https://suraj-singh-sde.github.io/portfolio