I am a Computer Science Engineer at NIT Karnataka (2022–2026) with hands-on experience at Amazon and NIT Karnataka's AI Research Lab. I design and ship production-grade systems at the intersection of data engineering, applied AI, and full-stack development.
My engineering philosophy centers on building systems that are observable, fault-tolerant, and cost-efficient — from real-time streaming pipelines and RAG-powered AI agents to cloud-native backends and analytical data warehouses.
I bring a product engineering mindset to every layer of the stack: I do not just build features, I instrument them, measure them, and iterate on them with data.
Open To
- 💼 Software Development Engineering (SDE)
- 📊 Data Engineering & Analytics
- 🤖 Applied AI / ML Engineering
- 🎓 Internship & New Grad Opportunities — 2026
- 🌏 India & Global Locations
Languages
Frontend
Backend & Databases
Cloud, DevOps & Tooling
| Domain | Proficiency | Details |
|---|---|---|
| Retrieval-Augmented Generation (RAG) | ███████████ Expert | Query expansion, RRF search, streaming, citation grounding |
| Generative AI & LLM Integration | ██████████ Expert | Conversational agents, memory chains, proactive insights |
| Machine Learning Engineering | █████████ Advanced | Feature pipelines, model evaluation, production serving |
| NLP & Semantic Search | █████████ Advanced | Embeddings, vector stores, hybrid retrieval |
| Data Engineering & ETL | ███████████ Expert | Spark, Kafka, Airflow, dbt, DuckDB, real-time pipelines |
| Analytics Engineering | ████████ Advanced | BigQuery, Looker Studio, Power BI, Tableau, A/B testing |
| Statistical Analysis | ████████ Advanced | Hypothesis testing, cohort analysis, funnel modelling |
⬡ Sentinel — API Key Management Platform
Enterprise-grade API key lifecycle management platform built for security, observability, and scale. Implements OAuth 2.0 authentication, real-time WebSocket dashboards, Redis-backed rate limiting, and full AWS cloud deployment with integrated monitoring.
| Attribute | Details |
|---|---|
| Stack | Node.js, Redis, WebSockets, OAuth 2.0, AWS (EC2, RDS, CloudWatch) |
| Scale | Multi-tenant, production-ready with horizontal scaling support |
| Performance | Sub-millisecond rate limiting via Redis; async event-driven architecture |
| Security | OAuth 2.0 token flows, scoped API key policies, audit trail logging |
| Impact | Eliminates API key sprawl; provides real-time usage visibility per consumer |
| Repository | github.com/saithrishadaggupati/sentinel |
Sentinel addresses a critical gap in developer tooling: secure, self-hosted API credential management with granular access control, revocation workflows, and live consumption analytics — without third-party lock-in.
⬡ Research Copilot — RAG Pipeline with Streaming & Citations
Production RAG system that transforms research workflows using query expansion, reciprocal rank fusion (RRF) search, real-time streaming responses, and inline citation grounding. Designed to reduce hallucinations while maximising retrieval precision.
| Attribute | Details |
|---|---|
| Stack | Python, LangChain, OpenAI, Pinecone, FastAPI, PostgreSQL |
| Scale | Multi-document corpus support with chunked vector indexing |
| Performance | Streaming tokens with < 1s time-to-first-byte via async FastAPI |
| Security | Document-level access control; prompt injection guardrails |
| Impact | Reduces manual research time; grounded citations eliminate hallucinations |
| Repository | github.com/saithrishadaggupati/research-copilot |
This system implements multi-query retrieval with RRF re-ranking, ensuring that no single query bias dominates the result set — a critical reliability property for research-grade AI tooling.
⬡ Food Delivery Backend — Geo-Aware Distributed System
High-performance food delivery backend with real-time geospatial queries, asynchronous task execution, and secure multi-role authentication — built for operational correctness under concurrent load.
| Attribute | Details |
|---|---|
| Stack | Django, PostgreSQL + PostGIS, Redis, Celery, JWT |
| Scale | Handles concurrent orders with async task queues via Celery workers |
| Performance | PostGIS spatial indexing for sub-10ms proximity queries |
| Security | JWT-based auth with role scoping (customer / restaurant / delivery) |
| Impact | Full delivery lifecycle management from order placement to geo-tracked fulfilment |
| Repository | github.com/saithrishadaggupati/food-delivery-backend |
PostGIS enables native geospatial queries for restaurant proximity ranking and delivery zone filtering — a systems-level decision that eliminates application-side distance computation at scale.
⬡ Stock Market Analytics — Modern ELT Data Stack
End-to-end analytical pipeline for stock market data using a fully modern ELT architecture. Combines DuckDB's columnar query engine with dbt transformation logic and dual BI layer publishing to Power BI and Tableau.
| Attribute | Details |
|---|---|
| Stack | Python, DuckDB, dbt, Power BI, Tableau, Statistical Testing |
| Scale | Processes multi-year OHLCV datasets with columnar query optimisation |
| Performance | DuckDB in-process execution eliminates network overhead vs traditional OLAP |
| Security | Immutable source layer; transformation lineage tracked via dbt DAG |
| Impact | Dual BI publishing; statistical significance testing embedded in pipeline |
| Repository | github.com/saithrishadaggupati/stock-analytics |
dbt models enforce a strict staging → intermediate → mart layer, making the transformation logic version-controlled, testable, and auditable by any stakeholder in the data org.
⬡ India AQI Data Pipeline — Automated ETL with Data Quality Monitoring
Automated ETL pipeline ingesting real-time India AQI data into MongoDB Atlas with embedded data quality gates, anomaly detection, and operational dashboards — built for unattended production execution.
| Attribute | Details |
|---|---|
| Stack | Python, MongoDB Atlas, Airflow, Data Quality Framework, Dashboarding |
| Scale | National-scale AQI ingestion across multiple monitoring stations |
| Performance | Scheduled pipeline with incremental loading and deduplication |
| Security | Schema validation and quality assertions before persistence |
| Impact | Enables longitudinal AQI trend analysis; alerts on data quality degradation |
| Repository | github.com/saithrishadaggupati/india-aqi-pipeline |
Data quality monitoring is first-class in this pipeline: each run computes completeness, validity, and freshness metrics — surfaced on the operational dashboard before any downstream consumption.
⬡ Finance Agent — Conversational Financial Analysis
LLM-powered finance agent with persistent memory, tool-augmented reasoning, and proactive insight generation. Enables natural-language querying of financial data with context retention across sessions.
| Attribute | Details |
|---|---|
| Stack | Python, LangChain, OpenAI, Financial APIs, Vector Memory |
| Scale | Session-persistent memory with long-context summarisation |
| Performance | Tool-augmented retrieval with parallel API fan-out |
| Security | Sandboxed tool execution; no raw financial data in prompt context |
| Impact | Replaces static dashboards with a proactive, queryable financial intelligence layer |
| Repository | github.com/saithrishadaggupati/finance-agent |
Memory architecture uses a hybrid short-term buffer and long-term vector store, enabling the agent to maintain coherent multi-session financial conversations without token explosion.
⬡ E-Commerce Funnel Analysis — BigQuery & Cohort Analytics
End-to-end e-commerce analytics system on BigQuery with funnel visualisation, cohort retention analysis, and A/B test significance reporting — built for product and growth team decision-making.
| Attribute | Details |
|---|---|
| Stack | BigQuery, Looker Studio, SQL, Cohort Analysis, A/B Testing |
| Scale | Large-scale event table analysis with partitioned BigQuery tables |
| Performance | Partitioned and clustered tables for cost-optimised analytical queries |
| Security | Row-level access policies per reporting persona |
| Impact | Identifies drop-off stages with statistical certainty; informs product roadmap |
| Repository | github.com/saithrishadaggupati/ecommerce-funnel |
A/B test results are evaluated using two-proportion z-tests with correction for multiple comparisons — ensuring product decisions are not made on noise.
⬡ SmartAttendance — QR-Based Role-Aware Attendance System
Secure, QR-driven attendance management platform with role-based access control for students, faculty, and administrators. Designed for institutional deployment with tamper-resistant session tokens.
| Attribute | Details |
|---|---|
| Stack | Python, Django, QR Code Engine, PostgreSQL, RBAC |
| Scale | Institution-wide deployment with multi-role concurrent access |
| Performance | QR token validation in < 100ms with session-bound expiry |
| Security | Time-bound QR tokens; role-scoped access; audit log per session |
| Impact | Eliminates proxy attendance; provides faculty with real-time attendance analytics |
| Repository | github.com/saithrishadaggupati/smart-attendance |
QR tokens are cryptographically signed and carry embedded session metadata — preventing replay attacks while maintaining a lightweight client-side scanning experience.
2024 · Remote / India
Worked within Amazon's BI engineering function, contributing to data pipeline reliability, metric framework development, and analytical reporting for internal stakeholders. Designed and maintained ETL workflows at scale, ensured data quality across reporting surfaces, and collaborated with cross-functional teams to deliver actionable intelligence from large structured datasets.
- Designed and optimised SQL-based ETL pipelines processing large-scale operational data
- Built and maintained BI dashboards surfacing KPIs to product and operations stakeholders
- Implemented automated data quality checks reducing reporting anomalies
- Contributed to metric framework standardisation across reporting domains
2023–2024 · Surathkal, India
Conducted applied AI research at NIT Karnataka's research lab, focused on NLP and information retrieval systems. Prototyped and evaluated retrieval-augmented pipelines, contributed to research on semantic search architectures, and supported publication-quality experimentation with reproducible ML workflows.
- Researched and prototyped RAG architectures for academic document retrieval
- Evaluated retrieval strategies including dense, sparse, and hybrid (RRF) approaches
- Built reproducible ML experiment pipelines with tracked metrics and evaluation harnesses
- Contributed to internal research reports and technical documentation
| Recognition | Details |
|---|---|
| 🏢 Amazon Intern | Selected for Business Intelligence Engineering internship at Amazon |
| 🔬 NIT Karnataka AI Research | AI Research Internship at a premier NIT research lab |
| 🎓 NIT Karnataka | Admitted to B.Tech CSE at a top-10 National Institute of Technology in India |
| 🧠 8 Production Projects | Shipped 8 end-to-end production-grade engineering projects across AI, data, and full-stack |
| 🌐 Open Source | Active GitHub presence with projects spanning AI agents, data pipelines, and backend systems |
| 📊 Full-Stack Analytics | End-to-end analytics engineering from raw ingestion to BI layer and statistical testing |
Certifications will be added here. Providers: AWS, Oracle, NPTEL, Cisco.
current_focus:
learning:
- Advanced distributed systems design patterns
- LLM fine-tuning and RLHF workflows
- Real-time feature stores for ML serving
building:
- Production RAG systems with hybrid retrieval and streaming
- Cloud-native data pipelines on AWS (Glue, Kinesis, Redshift)
- Agentic AI systems with memory and tool orchestration
exploring:
- Multi-agent frameworks (LangGraph, AutoGen)
- Iceberg / Delta Lake open table formats
- Vector database architectures at scale
open_to:
roles:
- Software Development Engineer (SDE)
- Data Engineer / Analytics Engineer
- Applied AI / ML Engineer
level: Internship & New Grad 2026
locations: India · Global · Remote
"Systems that scale are not built by accident — they are designed with intention, instrumented with care, and improved with data."