Nov 2024 – Oct 2025 · 11 mos · Kolkata, West Bengal, India · Hybrid

• 🚀 AI & Intelligent Systems | Gen AI Architect
• Here’s a snapshot of advanced AI-driven solutions I’ve recently architected—leveraging GPT-4o, modular agent design, and enterprise-grade integrations to solve real-world challenges:
• 🔹 AI-Driven Multi-Agent Ticket Resolution
• Built a scalable, intelligent agent ecosystem to streamline JIRA ticket handling using GPT-4o.
• ✅ Specialized agents for analysis, RAG-based knowledge retrieval, summarization, and code recommendation
• ✅ Full-stack delivery using FastAPI, Chainlit, and a custom JIRA CUD interface
• ✅ Integrated Agentic & Hybrid RAG to elevate contextual precision
• 🔹 ITOps Workflow Intelligence Agent
• Engineered an AI agent for L1 IT operations—handling sentiment analysis, classification, and priority-based routing.
• ✅ Cross-agentic design using CrewAI, LangGraph, and Atomic Agents
• ✅ Enabled Agent-to-Agent (A2A) collaboration and Multi-Channel Protocol (MCP) for dynamic support
• 🔹 PII Compliance & Privacy Audit Suite
• Led development of an AI-powered privacy enforcement pipeline across IT workflows.
• ✅ Real-time PII detection and redaction using Microsoft Presidio and Phi-3
• ✅ Developed a structured JSON-based PII scoring metric for standardized audits
• 🔹 Smart Mailbox – AI Based Invoice Processor Engine
• Architected a GPT-4o solution to extract structured data from invoice emails.
• ✅ Auto-ingests and processes invoices, pushing output to Microsoft Dataverse
• ✅ Securely stores files in Azure Blob (VPC-enclosed)
• ⚡ ~30 seconds avg. processing time per document
• 🔹 AI Piper – FDD Generator from Proposal Docs
• Created an LLM-powered system to transform PFCs into complete Functional Design Documents (FDDs).
• ✅ Combines OCR, visual parsing, and LLM-based content generation
• ✅ Integrated with Azure and SharePoint for seamless collaboration and formatting
• 📌 Passionate about building transformative AI solutions that reduce manual overhead, improve operational intelligence, and deliver business value at scale.

Mar 2023 – Nov 2024 · 1 yr 8 mos · Kolkata, West Bengal, India · Hybrid

• Currently engaged in the creation of a Gen AI Assistant System to help telecom operations in the daily activity for both structured data and unstructured data using large language model to extract best possible answer for a question asked from multisource documents.
• Fine Tuned code llama and llama-3 model for text to SQL generation task for some specific use cases along with various prompts.
• Harnessing speech-to-text and text-to-text translation models for automatic speech recognition and seamless translation tasks.
• I'm actively engaged in prompt engineering work, refining the system's ability to generate responses effectively.
• Detect anomalies occurring in Microwave nodes resulting degraded user experience using Autoencoder.
• Anticipating and preemptively identifying RRU faults.
• Implemented traffic balancing in Mobily's core network, in collaboration with Ericsson Managed Services, boosted operational autonomy during high-demand events. This led to an award-winning solution (AI Ops Award, FutureNet MENA 2024), part of the Ericsson Operations Engine. Using AI, we forecast and dynamically adjusted traffic across core nodes 12 hours in advance, enhancing efficiency, reliability, and user experience.
• Implemented custom NER strategy to recognize telecom specific vocabulary with the correct tag using Spacy and BERT based model.
• Developed a model using Roberta to identify valid question based on the telecom keywords using NER tag as a feature in the sequence classification model.
• Added domain specific keyword to tokenizer vocab to enhance contextual understanding of Roberta model.

Dec 2021 – Feb 2023 · 1 yr 2 mos · Kolkata, West Bengal, India · Hybrid

• Working for Tavant's cloud based analytics product called TMAP.
• Worked on warranty Analytics module , service contract analytics , IOT analytics module etc.
• Developed a supervised ML model to predict and discover factors impacting service contract price.
• Performed Reliability Analysis for the OEM and NON-OEM parts.
• Developed ML Model to capture suspicious claims in the manufacturing industry.
• Developed a Predictive Analytics Pipeline to forecast the telematics Sensor Reading from the past 3 months data for the next week using the Temporal Fusion Transformer algorithm and showcased the forecasting using the Power BI dashboard and saved the output in the Database. Later used the forecasted data to detect anomalies in the sensor reading for predictive maintenance purposes.
• Deployed the ML model to Azure ML and created an endpoint.
• Deployed the ML model to Azure as a Time Trigger Azure Function Service.
• Developed suspect claim analytics model to detect suspicious claims raised in the manufacturing industry.
• Developed AI based contextual search engine to get the relevant document in the search result according to the similarity score. Implement multiple advanced search feature like related search , trending search, automatic spell correction in search term, autocomplete and autosuggestion , Question-Answering .

Jun 2018 – Dec 2021 · 3 yrs 6 mos · Kolkata, West Bengal, India

• Developed a supervised ML model to predict and discover factors impacting credit default rate.
• Developed Regulatory Smart Search engines using Elasticsearch to get similar publications from regulations inventory by applying BERT text embedding and cosine similarities.
• Developed a sentiment analysis model with NLP to classify the sentiment of user feedback.
• Developed Question/Answering and Information Extraction model using Transformers framework.
• Developed an AI model to classify given banking regulation as applicable or not applicable to client business.
• Developed an AI model to tag Activity, Risk, and control description in GRC Library based on historical data.
• Implemented python program to extract entities from banking documents to automatically populate the credit application form.
• Developed ML pipeline to deploy AI model into a production environment using Flask web application framework, Cheroot and gunicorn web server, Docker containers, Jenkins CI/CD pipeline.

Mar 2014 – May 2018 · 4 yrs 2 mos · Kolkata, West Bengal, India

• Worked as Application enhancement and operation analyst for a USA Based Banking Firm.
• Automated manual tasks by writing Unix bash shell scripts.
• Good understanding of application deployment and workflow.
• Issue Resolution and Root Cause Analysis.
• Monitoring Batch jobs and scheduling them in Autosys.
• Providing support during code release, deployment, patching activities, and scheduled maintenance application outages.
• Documentation of the new application features.
• Involved in scripts enhancement and code fixes.
• Deep understanding of Unix Shell Scripting, Spring Batch console, Autosys, and handling service Now to track incidents.
• Data technologies - SQL
• In-depth understanding of enterprise automation tool - ITRS Geneos, AutoWAtch Plus, RT View.