Headline

Subject Matter Expert – GenAI | Forward Deployed Engineer | LLMs | AI/ML Solutions | Deep Learning Specialist | FastAPI | NLP | Computer Vision | Python | ML Deployment Architect | Scalable ML Pipelines | Open To Work

About

Senior Software Engineer specializing in Generative AI, Machine Learning Systems, and Scalable Backend Architecture, with 8 years of experience taking complex AI systems from research to high-throughput production. Currently at AB InBev, I architect stateful agentic AI platforms and intelligence engines using LangGraph, Redis, and WebSockets, deploying resilient multi-turn reasoning frameworks across web and iOS. My engineering background spans the full AI lifecycle—from training custom deep learning models to architecting production microservices and CI/CD pipelines. Key Career Highlights & Impact: - Agentic AI & Evaluation (AB InBev): Built "God Mode" for ABI's CompIntel app for automated financial data retrieval; created LLM evaluation frameworks covering tool selection, SQL generation, and reasoning across 15+ enterprise use cases. - Enterprise Modernization (Omnicom / Annalect): Transitioned 5 vendor dependencies into an internal 30+ member engineering org, cutting costs by 45%; architected DataVista, a GenAI analytics assistant that accelerated reporting by 50% and modernized 25+ legacy pipelines. - Supply Chain Automation (Annalect): Automated 80% of Amazon fee recovery, processing $600K+ monthly invoices at 28% higher accuracy and saving 150+ operational hours/month. - High-Throughput ML Serving (AdvaRisk): Architected an AI/ML platform integrating OCR, NER, and NLP across 7 production geographies with 90%+ accuracy; engineered FastAPI pipelines improving deployment efficiency by 72% and document processing speed by 80%. - Healthcare AI (Relevance Lab): Built deep learning OCR engines for CMS-1500 forms with 95% accuracy across 56+ global hospitals. Core Technical Stack: - GenAI: LangGraph, LangChain, RAG, Hugging Face, LlamaIndex, Agentic Workflows, LangFuse - Backend: Python, Java, SQL, FastAPI, Flask, Spring Boot, Microservices - ML / CV / NLP: PyTorch, TensorFlow, Scikit-Learn, OpenCV, TrOCR, Detectron2, spaCy, BERT - Cloud & MLOps: Azure, AWS, GCP, Docker, Kubernetes, CI/CD, PySpark, Snowflake, Redis, RabbitMQ Awards: Ace Performer Award, Noteworthy Key Performer Award, Relevance Lab Spot Award | Oracle Certified Java SE 8 Programmer. Open to conversations around GenAI engineering leadership and intelligent systems. Reach out at muzaffartasgoankar@gmail.com.

Location

Bengaluru

AI stack

claude code

claude

codex

cursor

deepseek

ollama

Work experience

Senior Software Engineer - Gen AI at AB InBev, Nov 2025 - Present · 1 yr 1 mo, Bengaluru, Karnataka, India

Subject Matter Expert - ML & Data at Omnicom Media Group, Jul 2024 - Oct 2025 · 1 yr 4 mos, Bengaluru, Karnataka, India, ● Successfully led a team of 15+ engineers through a strategic transition from 5 external vendors to in-house capabilities, improving operational efficiency and reducing costs. ● Strengthened 3 functional pods (Supply Chain, Retail, Operations Reporting), driving development efforts that scaled growth by 60%. ● Automated 80% of the supply chain fee recovery program for Amazon, encompassing shortages, price claims, chargebacks, and CoOp processes. ● Spearheaded modernization of 25+ legacy processes by implementing Python, data engineering, and machine learning solutions, enhancing efficiency and reliability. ● Building an automated CoOp pipeline, eliminating 150+ hours of manual work monthly and enabling efficient dispute management for 47+ clients involving $600K+ invoices.

Lead ML Engineer at AdvaRisk, Feb 2022 - Jun 2024 · 2 yrs 5 mos, Pune Division, Maharashtra, India, ● Engineered an automated table extraction algorithm to extract tables from scanned & machine-generated PDF documents for land documentation in 5+ states, improving accuracy by 40% and reducing processing time by 60%. ● Devised a cutting-edge Deep Learning-based Captcha model accurately deciphering captcha characters with 98% precision, streamlining scraping operations, and reducing manual effort for data extraction. ● Utilized FastAPI to develop APIs for serving Machine Learning models, ensuring seamless integration, and achieving end-to-end, which increased deployment proficiency by 72%. ● Developed an AI serving platform & pipeline for the seamless integration of 4 models:- NER (Named Entity Recognition), Transliteration, OCR & Name-matching along Natural language Processing. ● Implemented an efficient queuing mechanism using Celery and RabbitMQ to ingest land survey documents from government websites for faster document extraction with a speedup of 80%. ● Integrated custom APIs for ad-hoc tasks, ensuring the system's versatility in handling diverse requirements, and produced & monitored complex business logic for 12+ geographies. ● Led and assembled a team of ML Engineers to address complex business challenges related to land details and encumbrance information across 15 diverse geographical states. ● Spearheaded the application of AI methodologies to devise innovative solutions for unique business problems that served an accuracy of more than 90% for 7 production live geographies.

Senior Software Engineer - Data Science at Relevance Lab, May 2021 - Jan 2022 · 9 mos, Bangalore Urban, Karnataka, India, ● Collaborated with 7+ American & European multinational companies having diverse backgrounds like Healthcare-Pharma /wearables, Insurance, etc. which improved revenue growth rates by 30%. ● Expanded analytics & ERP flow by 62% with custom codes using PySpark technology & Azure systems for analytics. ● Introduced & mastered 6 APIs using efficient & reliable frameworks like Neo4j, Docker, Kubernetes, Bit bucket, Git, Nginx & Jenkins to meet market standards in terms of architecture which improved speed up by 54% overall.

Software Engineer - Data Science at Relevance Lab, Aug 2018 - May 2021 · 2 yrs 10 mos, Bangalore, ● Extracted unstructured scanned CMS (Centers for Medicare & Medicaid Services) forms of 56+ global hospitals by developing an OCR engine with 95% above accuracy with negligible errors for Inova clients using modules like Poppler, pdf2image, OpenCV, Tesseract, Keras CRNN, TensorFlow, Pillow, Scipy, EasyOCR, torch, Scikit-image, ImageMagick, etc. ● Formulated deep learning Classifier model to enable automated checkbox fields in the CMS forms. ● Analyzed 7+ complex data sets using fuzzy-wuzzy algorithms, to predict the actual result & prepared 30+Regex logics. ● Configured Linux 10+ servers (Redhat & CentOS) to host platform documents using Apache HTTP & Tomcat with Java. Along with the illustrated & customized Apache Solr search for the data screens over the SPECTRA platform for 20+ POCs to clients, which involved 6 top-notch brands like Hershey, Merck, Johnson & Johnson, etc. ● Ingested SAP data from 4 ADM database servers using SQLAlchemy & psycopg2 with international clients. And created 8+Java microservices APIs for CRUD operations(data retrieval & manipulation) using SpringBoot technology.

Education

Bachelor's degree, Computer Science, Shri Dharmasthala Manjunatheshwara College of Engineering and Technology (SDMCET)

PUC, Science, Vidyaniketan P.U. Science College

Public links

github: https://github.com/Mohammedmuzzu

youtube.com: https://youtube.com/@muzaffartasgoankar