Hi, I'm Saif Raza Khan
Software Developer (GRA), NASA-IMPACT | ex-R&D Network Engineer
Helping build NASA's Science Discovery Engine at IMPACT through search pipelines, vectorization, and scalable AWS infrastructure. Previously engineered carrier-grade telecom networks, now applying that systems mindset to software. Passionate about clean code, reliability, and solving complex technical challenges.
About Me
I'm a Graduate Research Assistant with NASA's Interagency Implementation and Advanced Concepts Team (IMPACT), where I help build the Science Discovery Engine (SDE). My work spans search pipelines, embeddings, ML classifiers, and scalable AWS infrastructure - all focused on advancing Open Science by making NASA's scientific knowledge more discoverable and usable for researchers worldwide. I began as a network engineer working on carrier-grade telecom systems. I've come to strongly believe that everything is f(time) - with patience, persistence and effort, we can figure things out. I don't claim to be an expert - just an eager learner, a logic lover and someone who likes to ask "why not?" as often as "how?". While networking and software development are my comfort zones, I believe modern engineers should hold a broad, high-level map across domains. Outside of code, I'm equally fascinated by society, liberty, and philosophy - conversations about the human condition inspire me as much as system design docs. I like questioning existence, debating purpose and spotting patterns across both life and technology. No matter the project, I approach it with curiosity, humility and an "all-in" mindset - because whether it's science, software or society, I love adding value wherever I can. And if there's some humor along the way? Even better.
Education
M.S. in Computer Science
University of Alabama in Huntsville
Huntsville, Alabama • Aug 2024 – Present
GPA: 4.0/4.0
B.Tech in Electronics and Communication Engineering
University Visvesvaraya College of Engineering
Bangalore, India • Jun 2019 – Jun 2023
GPA: 8.81/10
Technical Skills
Programming & Backend
- Languages: Python, C, C++; Bash (Working knowledge: Go)
- Backend & APIs: Flask, FastAPI; REST (OpenAPI/Swagger), pagination, idempotency, error contracts, versioning
- Quality & Testing: pytest, fixtures/mocks, coverage targets; ruff/black, mypy
Search, Data & ML
- Search / IR: OpenSearch Serverless (HNSW), schemas/mappings, hybrid semantic+keyword, reranking, ingestion/indexing pipelines
- ML / Vectorization: SageMaker endpoints & processing, external mean-pooling, (experiments) binary quantization
- Data & Storage: PostgreSQL/MySQL (schemas, indexes), S3; Redis (cache-aside, TTL/evictions)
- Messaging & Batch: Kafka; retry/backoff patterns, DLQ semantics
Cloud & Infrastructure
- AWS: EC2, SageMaker, OpenSearch Serverless, IAM, CloudWatch (metrics/logs/alerts)
- Containers & Orchestration: Docker, docker-compose; ECS/Fargate; basics of EKS
- Event/Data Pipelines: SQS, SNS, Kinesis, Step Functions
- IaC: Terraform (basics)
Systems & Performance
- Data Structures and Algorithms (DSA)
- Distributed Systems: consistency & idempotency, backpressure, retries, failure handling, pagination/limits, schema evolution
- Performance & Load: profiling (cProfile/py-spy), load testing (k6/Locust)
- Observability: structured logging, latency/error dashboards, ops metrics & runbooks (Working knowledge: OpenTelemetry, SLO-driven alarms)
- Security: OAuth2/JWT; AWS KMS/Secrets Manager
Networking (Prior Specialty)
- Protocols: BGP, OSPF, MPLS, ERPS, VLANs, IPv4/IPv6; LLDP/MEPs; DHCP/DNS
- Diagnostics: Wireshark, Ixia, Spirent; SNMP, NetFlow/sFlow/telemetry
- Embedded/Linux: firmware upgrades, daemon/process debugging, fault-injection & recovery validation
Tooling & Collaboration
- Git/GitHub workflows, PR reviews, Issues/Projects
- CI/CD: GitHub Actions (build/test), artifacts
- Documentation: markdown docs, architecture diagrams, reproducible READMEs
- Linux-first workflows, shell scripting, GNU toolchain
Research
Trust at Speed: Measuring & Controlling LLM Faithfulness Under ANN Retrieval Approximations in RAG (Masters Thesis): This thesis moves beyond generic RAG evaluation to provide empirical, causal evidence of how approximate nearest neighbor (ANN) retrieval affects LLM trustworthiness. It examines how retrieval knobs (efSearch, nprobe, k) shape coverage, latency, and hallucination rates across FAISS, Qdrant, and OpenSearch.
Making BGP over QUIC Real: Empirical Evidence for Performance, Interop, and Adoption (Independent Research): Investigating the real-world performance and deployment challenges of BGP over QUIC beyond basic feasibility, focusing on convergence times, head-of-line blocking reduction, and route burst handling under realistic Internet conditions.
Professional Experience
Software Developer (GRA)
NASA-IMPACT
- Contribute to NASA's Science Discovery Engine (SDE) by developing search and data software systems
- Design OpenSearch Serverless schemas and vector search, ingestion/indexing pipelines, and SageMaker-based vectorization
- Implement authentication hardening and CI/CD with GitHub Actions
- Prototype IR stacks (e.g., Meilisearch) and evaluate NASA data sources (e.g., NTRS) for retrieval quality
R&D Network Engineer
Tejas Networks Ltd.
- Diagnosed complex network issues in Layer 2/3 switching, BGP, OSPF routing using Wireshark, Ixia, Spirent
- Designed and validated ERPS with sub-50ms failover, MPLS tunnels, and carrier-grade reliability
- Performed system-level debugging on embedded telecom systems (TJ1400/TJ1600)
- Conducted large-scale traffic simulations (1G–100G) for backbone resilience validation
Intern
Tejas Networks Ltd.
- Completed comprehensive telecom networking fundamentals training program
- Earned 11 Tejas Networks certified courses covering various networking technologies
- Specialized in ERPS (Ethernet Ring Protection Switching) protocol implementation
- Presented technical project on network resilience to stakeholders and engineering teams
Featured Projects
🔌 Ethernet Ring Protection Switching (ERPS)
Implemented carrier-grade network resilience with sub-50ms failover mechanisms for telecom infrastructure.
Beyond the Code
While I'm passionate about technology and engineering, I believe the most interesting solutions come from drawing connections across diverse fields of knowledge and experience.
Philosophy & Ideas
Exploring fundamental questions about existence, consciousness, and the nature of reality.
- Ancient Stoic philosophy and its modern applications
- Philosophy of mind and consciousness studies
- Ethics in technology and artificial intelligence
- Eastern philosophical traditions and meditation practices
Mixed Martial Arts
Training in various combat disciplines, focusing on technique, discipline, and mental fortitude.
- Brazilian Jiu-Jitsu for ground game and problem-solving
- Boxing for striking fundamentals and footwork
- Muay Thai for clinch work and conditioning
- Mental discipline and stress management through combat sports
Indian Classical Music
Deep appreciation for the mathematical precision and emotional depth of Indian classical traditions.
- Raga structures and their mathematical foundations
- Tala systems and rhythmic complexity
- Improvisation within traditional frameworks
- The spiritual and philosophical aspects of classical music
Continuous Learning
Voracious reader with interests spanning science, philosophy, history, and human nature.
- Biographies of great scientists and thinkers
- History of technology and scientific revolutions
- Cognitive science and human psychology
- Science fiction as a lens for exploring future possibilities
Reflections & Perspectives
“The more we know, the more we realize how vast the unknown is.”— On knowledge and humility
“Patience with the process is often the fastest path to real understanding.”— On learning and growth
“In the end, wisdom is less about being right, and more about seeing clearly.”— On wisdom and perspective
“Seek knowledge from the cradle to the grave.”— On lifelong learning