Headline

Senior Software Engineer @ CriticalRiver | Python · FastAPI · Go · GenAI (LLMs, RAG, LangChain) | AWS · Azure · GCP | Ex-Kraft Heinz, Dunzo, TCS

About

Senior software engineer with 5+ years building cloud-native backend and full-stack products in Python (FastAPI, Django), Go, Node.js, and React across AWS, Azure, and GCP. Currently at CriticalRiver building production GenAI applications — LLM workflow automation on OpenAI/Azure OpenAI and Anthropic Claude, with RAG pipelines (LangChain, pgvector, Elasticsearch, FAISS/Chroma) that cut manual support triage by 30%. Highlights: - CriticalRiver (Ministry of Power)— Revenue Leakage Detection platform, showcased at the MoP National Conference - Kraft Heinz — led architecture of the global NPS platform on Azure; mentored 2 engineers; cut analytics refresh time 40% for 150+ users - Dunzo — cut latency 30% on Home service handling 1K+ req/s; AOV +₹25/order - TCS — CloudWatch Synthetics monitoring, MTTD −60% Reliability record: 99.9% uptime, distributed tracing (Elastic APM), Kafka, Docker, Kubernetes, CI/CD. Startup co-founder (GUSEC-incubated).

Location

Bengaluru

AI stack

claude code

claude

codex

opencode

Work experience

Senior Software Engineer at CriticalRiver Inc., Nov 2025 - Present · 10 mos, Remote, Building production GenAI and analytics platforms for enterprise and government clients. - Architected and delivered a Revenue Leakage Detection platform for India's Ministry of Power (FastAPI + React), translating ML-driven leakage indicators and KPIs into actionable insights — selected for showcase at the MoP National Conference (Dec 2025). - Built an LLM-powered Service Management Assist application on OpenAI/Azure OpenAI and Anthropic Claude with a LangChain RAG pipeline (pgvector, Elasticsearch, FAISS/Chroma), automating 10+ support workflows and reducing manual triage effort by 30%. - Delivered the Leaflinks analytics platform — 3 production dashboards (React + Django) powered by AWS Lambda ETL pipelines automating data ingestion and transformation. - Own production deployment and release automation end-to-end with Docker and Nginx on AWS, establishing release practices that sustain 99.9% uptime.

Senior Software Engineer at Kraft Heinz, Jan 2024 - Oct 2025 · 1 yr 10 mos, - Developed and enhanced a high-end data visualization platform, enabling machine learning insights for decision-makers through an interactive Angular web application. - Spearheaded the NPS Project within Kraft Heinz, mentoring 2 engineers and collaborating with stakeholders to drive customer satisfaction improvements. - Winner of the Commercial Insights Challenge, showcasing strategic thinking and analytical expertise by extracting business insights from complex datasets. - First hire in the organisation in Full Stack Development role, setting a precedent for future projects and driving significant impact in Enterprise Data & Analytics.

Software Engineer - 2 at Dunzo, Aug 2022 - Dec 2023 · 1 yr 5 mos, Remote, - Led the Custom Pages Project, optimizing data flow and consumer behavior insights, increasing Average Order Value (AOV) by ₹25. - Implemented BxGy (Buy X Get Y% Off) initiative, contributing to an ₹8 boost in AOV and driving revenue growth. - Resolved critical memory leak issues using the pprof profiler, enhancing system reliability and operational efficiency. - Collaborated across product, marketing, and design teams for the Category Page Redesign, improving user engagement and driving a ₹15 increase in AOV. - Mentored junior developers, fostering a culture of problem-solving and knowledge sharing.

Systems Engineer at Tata Consultancy Services, Sep 2020 - Jul 2022 · 1 yr 11 mos, Ahmedabad, - Built multifarious AWS lambda functions communicating with S3 and DynamoDB and hosting API Gateway endpoint in front of it. Integration of the API Gateway endpoints in the Node backend. - Restructured the legacy parsers of JDX, XLSX, and CSV files in Python , reduced the number of function call to compute the length of dataset and computing the length efficiently, resulting in increase of parsers efficiency by 35%. - Built export functionality using rxjs in Node and material-ui in React for the legacy applications data table. - Restructured the default drop-down to allow searchable select functionality in a Node and React application. - Ideation and implementation of monitoring multiple clients applications endpoint using the Cloudwatch Synthetics Canary.

Education

Bachelor of Engineering - BE, Information Technology, L.D. College of Engineering