Shubham Ojha
AI Engineer & Software Architect
Software Engineer with 8+ years of experience in AI/ML engineering, backend systems, scalable cloud-native architecture, and GenAI product development. Specializing in building production-ready LLM-powered applications, RAG systems, and enterprise AI solutions using AWS, Azure, Kubernetes, and FastAPI.
Professional Experience
AI Consultant
Independent
Nov 2025 - Present • India | Multiple Clients
- ▸Consulting for companies on LLM integrations, GenAI product development, and scalable backend architectures
- ▸Designing multi-agent systems that take a user query → break into tasks → execute via code, search, and API calls
- ▸Building healthcare-focused LLM assistants: grounded answers, citations, dosage lookups, and document interpretation
- ▸Creating document intelligence systems for medical and legal PDFs (classification, summarization, RAG)
- ▸Designing and deploying GenAI-based mobile and web apps (React Native, FastAPI, Azure/AWS LLM pipelines)
- ▸Architecting serverless LLM systems using Azure Web Apps + SSE, Azure Cognitive Search, and FastAPI
- ▸Supporting startups with AI strategy, architecture, prototyping, and cost-optimized inference setups
Technical Lead
Space Inventive
Oct 2023 - Oct 2025 • Bengaluru
- ▸Led architecture for an enterprise-scale M&A document intelligence platform handling 1–5 TB per deal
- ▸Designed GDPR-compliant pipelines using AWS SageMaker, Kubernetes, and OpenSearch
- ▸Built an OCR + NER + extraction pipeline using Lambda, API Gateway, ECS/K8s components
- ▸Built an LLM-based content classification platform, reducing labeling time by 50%
- ▸Created a RAG-powered sales intelligence tool on Azure Cognitive Search + OpenAI, cutting pitch prep time by 40%
- ▸Developed an AI Interview Simulator using FastAPI + LLMs, doubling practice throughput
- ▸Built an AI presenter with real-time TTS/STT + LLM Q&A for live product demos
- ▸Fine-tuned a medical chatbot using QLoRA with improved clinical answer accuracy
- ▸Implemented MLOps pipelines for CI/CD, monitoring, model updates & drift reduction (35%)
- ▸Set up 99.9% uptime Azure infra with containerized deployments and observability
Senior Software Engineer
AI Palette
Jan 2023 - May 2023 • Bengaluru
- ▸Optimized Elasticsearch + Django-based analytics (30% lower latency)
- ▸Automated ETL pipelines → 5× throughput, 80% manual work eliminated
- ▸Improved search stack with autocomplete, spell correction, and translation for global datasets
- ▸Built React + D3 dashboards for real-time insights
- ▸Reduced Python bulk-computation time by 20%
Project Engineering Lead
Adcuratio Media
Mar 2020 - Dec 2022 • Bengaluru
- ▸Migrated monolith → Dockerized microservices, improving deployments by 50%
- ▸Standardized data models and APIs, reducing integration issues by 40%
- ▸Led three engineering squads, bridging tech + business
- ▸Implemented real-time Elasticsearch pipelines (reduced batch lag from 2 hours → 5 minutes)
- ▸Built an A/B testing framework, improving ad conversions by 22%
- ▸Ensured 99.9% uptime with HA and redundancy strategies
Software Engineer
Adcuratio Media
Feb 2019 - Mar 2020 • Bengaluru
- ▸Refactored core modules (35% lower complexity) using SOLID principles
- ▸Built REST APIs for ad targeting (99% targeting accuracy)
- ▸Containerized services → zero-downtime deployments
- ▸Built real-time alerts using Django + SendGrid
- ▸Created ad optimization models (15% CTR improvement)
Assistant System Engineer
Tata Consultancy Services
Dec 2017 - Feb 2019 • Bengaluru
- ▸Automated DN provisioning (reduced manual effort by 90%)
- ▸Audited 2,000+ servers for logging compliance
- ▸Handled multi-region production incidents (30% MTTR reduction)
Featured Project
Drelo.in — AI-Powered Health Companion App
React Native + FastAPI + GenAI
Built a mobile health assistant app that lets users chat with an AI to understand symptoms, ask health questions, and summarize medical reports. Integrated LLMs with medical prompting for safe, structured responses.
Key Features:
- ✓Symptom gathering and smart health flows
- ✓Summary generation from medical reports
- ✓Suggested follow-up questions for better clarity
- ✓Extraction from uploaded medical files
- ✓Backend built with FastAPI, supporting SSE-based streaming responses
- ✓Private health data handling with token-based access
Technology Stack:
React NativeFastAPIAzure OpenAIEmbeddingsSSE
Technical Skills
Programming & Frameworks
PythonSQLJavaScriptReactReact NativeFastAPIDjangoBashYAMLStreamlitD3.js
AI / ML / GenAI
NLPComputer VisionDeep LearningLLMsGenerative AIPrompt EngineeringRAG SystemsEmbeddings (CLIP, BERT, Ada)Model Fine-tuning (QLoRA)Classification & SummarizationMulti-agent architecture
Frameworks & Libraries
PyTorchscikit-learnTransformersLangChainLlamaIndexPydantic AISpaCyPandasNumPyMatplotlib
Cloud & DevOps
AWS (EC2, Lambda, SageMaker)Azure (Cognitive Search, Cosmos DB)DockerKubernetesNGINXCI/CDAzure DevOpsGitHub ActionsMLOps
Databases & Data Systems
PostgreSQLMongoDBCosmos DBElasticsearchRedisOpenSearchVector search
Engineering Practices
System DesignMicroservicesAPI ArchitecturePerformance OptimizationSOLID PrinciplesClean ArchitectureAgile/ScrumCode Reviews
Education
Data Science Nanodegree
Udacity
2019
B.Tech in Information Technology
Dr. A.P.J. Abdul Kalam Technical University
2013 - 2017 • 78.06%