Mitra Brinda Mukherjee
Building production AI systems — from dialogue-state chatbots and multilingual voice agents to deepfake detection and semantic search.
Building AI that matters.
I'm an AI Engineer focused on conversational AI, voice agents, and applied machine learning — shipping systems that work in production, not just in notebooks.
Currently a Junior Software Development Engineer (AI) at Steora Systems in Kolkata, I design and deploy dialogue-state-managed chatbots, Retell and Anam voice and video agents, and FastAPI backends on AWS and Azure, wiring them into CRMs, email automation, and third-party APIs.
My research is in ML for cybersecurity. As a Research Intern at Texas A&M University under Prof. Marcus Botacin (May–Jul 2025, College Station), and I am still working on it remotely. I built AutoSteamYARA — an AI-driven system for automated YARA rule generation. I used the Apriori algorithm to extract detection patterns from malware binaries and security telemetry, then implemented adaptive stream-learning models (adaptive Random Forest via the River library) so malware pattern detection could update continuously instead of relying on static batch retraining. I also designed automated validation and benchmarking pipelines for reproducible experiments, reliability analysis, and failure-case analysis.
I care about building systems that are not just accurate, but interpretable, responsible, and deployable.
I'm active on LinkedIn — if you want to talk AI, research, or opportunities, come connect with me there.
Career Timeline
My professional journey and experiences.
Junior Software Development Engineer (AI)
Designed and deployed dialogue-state-managed AI chatbots for business automation; built voice agents on Retell AI and video agents on Anam AI. Built FastAPI/PostgreSQL backend services on AWS and Azure, and integrated CRM systems, email automation, and third-party APIs into AI-powered platforms.
Research Intern, ML for Cybersecurity
Built AutoSteamYARA for automated YARA rule generation using Apriori pattern extraction from malware binaries. Implemented adaptive stream-learning models (River) for continuously updating malware detection, and designed validation and benchmarking pipelines for reproducible experimentation.
Core Stack
Technologies and focus areas I work with day-to-day.
Python
Primary language for ML pipelines, agents, and backend services
AI / ML
PyTorch, Scikit-Learn, computer vision, NLP, stream learning (River)
Conversational AI
Dialogue state management, LiveKit Agents, Retell AI, Anam AI, Sarvam AI
Backend
FastAPI, Flask, REST APIs, PostgreSQL, Supabase
Semantic Search
Sentence-BERT embeddings, vector search, hybrid retrieval pipelines
Cloud & Tools
AWS, Azure, Docker, Hugging Face, Git/GitHub
Let's connect!
Whether you have a specific project in mind, want to discuss machine learning architectures, or just want to network, my inbox is always open.