Senior Software Engineer · Toast
Anish Mahapatra
I build reliable AI platforms and production machine learning systems — the deployment workflows, model serving, and data infrastructure that sit between a model and a dependable product.
About
I'm a software engineer with 8+ years across AI, machine learning, and data platforms, working on production deployments for enterprise and Fortune 500 environments. Before Toast, I worked on AIOps platforms at SAP, LLM-driven analytics and large-scale data pipelines at 7-Eleven, and predictive systems as a Lead Data Scientist at Mu Sigma.
I care about making system boundaries and failure modes visible, treating reproducibility and observability as requirements rather than afterthoughts, and choosing infrastructure that matches actual scale rather than anticipated scale.
MSc in Data Science (Distinction) from Liverpool John Moores University, a postgraduate diploma in Data Science from IIIT Bangalore, and a BTech in Information Technology from Manipal Institute of Technology.
Experience
- 2026 — Present
Senior Software Engineer · Toast
Recently joined to work on production software systems.
- 2024 — 2026
Senior MLOps Engineer, AIOps Platform Lead · SAP
Led productionization of an enterprise AIOps platform for Fortune 500 clients on Kubernetes; built MLOps/LLMOps workflows with GitHub Actions, ArgoCD, and MLflow, and NLP-based data anonymization pipelines.
- 2022 — 2024
Senior AI & Machine Learning Engineer · 7-Eleven
Built LLM-driven enterprise analytics with LangChain and real-time data pipelines on Kafka, Spark, and Databricks; led migration of large-scale tables to Delta Lake and Power BI.
- 2018 — 2022
Lead Data Scientist · Mu Sigma
Progressed from Data Scientist to Lead Data Scientist over four years, leading ML pipelines and data engineering for Fortune 300 clients across demand forecasting and anomaly detection.
Selected work
A packaged PyTorch workflow with leakage-safe data splits, reproducible training, FastAPI serving, tests, CI, Docker, and Kubernetes manifests.
A deployed Next.js application for AI system design and interview preparation, with a server-side contact workflow and Supabase persistence.
A local voice AI loop for macOS combining speech recognition, an LLM, speech synthesis, turn history, and latency tracing.
An architecture blueprint for payment event contracts, failure classification, retry decisions, evaluation, and production monitoring.
A Streamlit RAG prototype for querying PDF, DOCX, PPTX, and XLSX files using FAISS and LangChain.
Writing
- Designing an End-to-End AI Platform for Payments and Retail (AWS) Apr 2026
- Installing Claude Code for Agentic Engineering Feb 2026
- How to Set Up an Enterprise-Ready Production Environment for Data Engineering Jul 2024
- What is LangChain? (AI Crash Course) Feb 2024
- Build a CI/CD Pipeline on AWS Nov 2023
Technical interests
Production ML and LLM serving, evaluation and observability, Kubernetes-native delivery, and data contracts for real-time and batch pipelines.