GenAI8 min read
Building Production RAG Systems: Lessons from Enterprise Deployments
Key architectural decisions, retrieval optimization, and evaluation frameworks for RAG at scale.
Jan 15, 2025Read more
Open to new opportunities
Hi, I'm Sashank
ML Engineer
Building production ML systems that deliver measurable impact. Specializing in GenAI applications, RAG architectures, and computer vision pipelines. MS in Computer Science, Boston University.
3+
Years Experience
5
Production Systems
$500K+
Value Delivered
AWS 2x
Certified
Scroll
Experience
Building production ML systems with measurable business impact
Full-time
Machine Learning Engineer
HTC Global Services
Jan 2024 – PresentRemote
95%First-Query Resolution40%Discovery Time Reduction$100K/yrQA Cost Savings
- ›Designed and validated a multimodal RAG system (pgvector + LangChain) for natural-language document discovery across 1,000+ documents, achieving 95% first-query resolution
- ›Architected a GenAI call-center QA auditor (WhisperX + GPT-4o + DSPy) with human-in-the-loop scoring, scaled to 10K+ calls/month
- ›Developed a statistical anomaly detection pipeline for NPDS poison control data on 9M+ records in AWS SageMaker
- ›Led development of a custom document field-detection and extraction pipeline for 2M+ historical genealogy records using Faster R-CNN
- ›Mentored 3 interns on evaluation design and reproducible experimentation
GPT-4oWhisperXDSPyLangChainpgvectorFaster R-CNNAWS SageMakerPythonPostgreSQL
Intern
Machine Learning Engineer Intern
HTC Global Services
Sep 2023 – Dec 2023Remote
15%Shrinkage Reduction$200K/yrLoss Prevention50%Labeling Throughput
Intern
Data Science Intern
Technocolabs Softwares
May 2021 – Jul 2021Remote
86%Model Accuracy35%Processing Time Saved$200KDefaults Prevented
Projects
Production systems and research that drive real-world impact
GenAIOct 2025 – Nov 2025
Multi-Tenant GenAI Schema Translation Agent
Converts natural-language requests to canonical SQL then tenant-specific SQL across 6+ heterogeneous schemas
98%Successful Execution95%Latency Reduction6+Schemas Supported
- Self-correcting validation loop with schema-aware column extraction and retry/regeneration
- Deterministic field-mapping cache + fast-path rewrites for 95% latency reduction
- Prevents hallucinated columns/joins from reaching execution
GPT-4oDSPyLangChainPostgreSQLPython
GenAI2024
Multimodal RAG Document Discovery
Enterprise RAG system for natural-language discovery across 1,000+ documents with 95% first-query resolution
95%First-Query Resolution40%Discovery Time Reduced1000+Documents Indexed
- Gold-question set for tracking retrieval success
- Multimodal embeddings for documents with mixed content
- Deployed across 10+ teams at enterprise scale
pgvectorLangChainOpenAIPostgreSQLPython
MLOps
Statistical Anomaly Detection Pipeline
Detects emerging substance signals 3 weeks early from 9M+ NPDS poison control records
9M+ Records Processed3 weeks Early Signal Detection
CV
Document Field Detection & Extraction
Faster R-CNN pipeline for key entity localization in 2M+ historical genealogy records
~90% Bounding-Box Accuracy40% Manual Review Reduced
ML
High Entropy Alloys Property Prediction
ML model predicting mechanical properties of high entropy alloys with R² = 0.931
0.931 R² Score0.029 RMSLE
Skills
Technical expertise across the ML stack
All Technologies
PythonRAG SystemsPandasLangChainscikit-learnLangGraphPrompt EngineeringYOLOv8NumPySQLDSPyFaster R-CNNPyTorchAWS SageMakerWhisperXDetectron2OpenCVXGBoostTime SeriesDockerpgvectorstatsmodelsMongoDBDatabricksAzure MLRJavaScriptJava
Writing
Technical deep-dives on ML engineering and GenAI
GenAI6 min read
DSPy in Production: From Manual Prompts to Optimized Pipelines
How we improved Cohen's κ from 0.55 to 0.85 using DSPy for automated prompt optimization.
Nov 20, 2024Read more
MLOps10 min read
Statistical Anomaly Detection at Scale: NPDS Surveillance Case Study
Building seasonality-aware baselines for 9M+ records and detecting signals 3 weeks early.
Sep 10, 2024Read more
Education
Boston University
Master of Science in Computer Science
GPA: 3.97/4.00January 2024
Experimental DesignMachine LearningDeep LearningNLP
NIT Warangal
B.Tech in Metallurgical & Materials Engineering
GPA: 3.50/4.00July 2018 – May 2022
Institute Merit ScholarshipTop 5% of studentsResearch on High Entropy Alloys
Get In Touch
Open to ML Engineer, GenAI Engineer, and Research Engineer roles. Let's build something impactful together.