Gaurav Surtani
Founding AI Engineer | Growth Engineering | Full-Stack ML Systems
Full-Stack AI Engineer specializing in building intelligent systems and AI-powered solutions. Experienced in developing end-to-end AI applications. Passionate about creating impactful technology that combines cutting-edge AI with robust software engineering practices.
...
Projects
...
Organizations
...
Years Exp
About Me
I'm a passionate Full-Stack AI Developer from San Francisco Bay Area, currently working with leading organizations including DeepLearning.AI. My journey spans from building intelligent solutions that solve real-world problems to contributing to cutting-edge research in space exploration and AI.
Current Focus
Agentic AI & LLMs
Speech-to-Speech Translation
Deep Research Tools
Full Stack Dev
ML Applications
Agentic AI & LLMs
Speech-to-Speech Translation
Deep Research Tools
Full Stack Dev
ML Applications
Key Interests
AI/MLProduct DevelopmentRapid PrototypingGenAIAgentic AINLPKnowledge GraphsIntegrationFull Stack DevelopmentData EngineeringGeospatial AnalysisRecommendation SystemsTime Series AnalysisHealthcare Analytics
Technical Skills
AI & Machine Learning
PythonTensorFlowPyTorchNLPKnowledge GraphsAgentic AIScikit-learnGenAILLM Agents (LangChain)YOLO v8Computer Vision
Data Engineering
PandasNumPyData VisualizationETL PipelinesStatistical AnalysisSQLMongoDBNeo4jBigQueryPlotly
Web Development
JavaScriptTypeScriptReactNext.jsNode.jsAngularFlaskTailwind CSSChrome ExtensionsREST APIs
DevOps & Cloud
DockerKubernetesAWSAzureGCPCI/CDGitMicroservices
Professional Experience
Founding AI Engineer
DeepLearning.AI
San Francisco Bay Area
September 2025 - Present
DevelopmentTechnical Stack
Claude APITavilyOpenRouterVector DBLangChainReactPythonFlaskMulti-Agent OrchestrationSemantic SearchVector EmbeddingsData Pipelines
Key Achievements
Building AI-native recruiting platform for open-world candidate sourcing [Claude API, Tavily, OpenRouter]Architecting data pipelines across multiple sources (LinkedIn, GitHub, company databases) from scratch [Python, Flask]Developing intelligent search with vector embeddings and semantic ranking [Vector DB, LangChain]Implementing multi-agent orchestration for deep candidate research and profile enrichment [Claude API, OpenRouter]Delivering full-stack application to pilot recruiters; iterating on product based on direct user feedback [React]
AI Recruiting Platform
AI-native recruiting platform for open-world candidate sourcing with multi-model orchestration
- •Architected data pipelines across multiple sources (LinkedIn, GitHub, company databases) from scratch
- •Developed intelligent search with vector embeddings and semantic ranking to identify high-value companies and surface qualified candidates
- •Implemented multi-agent orchestration (Claude, Tavily, OpenRouter) for deep candidate research and profile enrichment
- •Delivered full-stack application to pilot recruiters; iterating on product based on direct user feedback
Full Stack AI Engineer, Growth
Tassel
United States
June 2025 - September 2025
DevelopmentTechnical Stack
AzureVoice AITTSFFmpegPythonVideo ProcessingInfrastructureCost Optimization
Key Achievements
Built video pipeline that ingests graduation ceremonies, segments by student, and delivers personalized clips to 2M+ students — $8M+ revenue opportunity [FFmpeg, Python]Developed AI pronunciation system to fix TTS accuracy for student name announcements across Voice AI applications [Voice AI, TTS]Architected Azure infrastructure; migrated always-on to on-demand scaling for heavy pipelines — cut costs 40%, processing 8x faster [Azure]
Graduation Video Pipeline
AI-powered video pipeline for personalized graduation ceremony clips at scale
- •Built video pipeline that ingests graduation ceremonies, segments by student, and delivers personalized clips to 2M+ students
- •Developed AI pronunciation system to fix TTS accuracy for student name announcements
- •Architected Azure infrastructure; migrated always-on to on-demand scaling — cut costs 40%, processing 8x faster
- •$8M+ revenue opportunity from personalized video delivery
Full Stack AI Engineer, Growth
Bexorg, Inc.
New Haven, Connecticut, United States
May 2024 - May 2025
DevelopmentTechnical Stack
ReactNode.jsPythonFastAPILangChainGPTNext.jsTypeScriptPlotlyD3.jsNeo4jBigQueryDockerKubernetesREST APIsGraphQLTailwind CSS
Key Achievements
Solved critical pipeline bottlenecks: 10x throughput improvement enabling new enterprise customer acquisition [React, Node.js, Python]Built GenAI analytics platform with NL querying — key differentiator for sales; 99.9% uptime converted pilots to contracts [FastAPI, LangChain, GPT]Applied AI/ML/NLP expertise: anomaly detection, NLP extraction, CV for medical imaging [ChromaDB, Gemini API]Implemented graph-based data layer with GNNs, reducing query times 5x for real-time analytics [Neo4j, BigQuery]
Interactive Data Viewer
Real-time data visualization platform using Plotly, Flask, Neo4j Aura, and BigQuery
- •Engineered scalable data pipeline with graph visualization (vis.js), enabling real-time exploration of complex relationships
- •Architected unified data access layer reducing retrieval time by 45% through optimized graph algorithms
- •Designed interactive dashboards with real-time updates, improving data analysis efficiency by 60%
- •Implemented role-based access control and audit logging, ensuring secure access for 40+ users across departments
Team Collaboration Calendar
Async communication and experiment scheduling system
- •Developed async-first calendar system reducing experiment deployment time by 85% (7 days → 3-4 hours)
- •Built automated scheduling engine reducing 1:1 meetings by 60% while maintaining team alignment
- •Created real-time experiment tracking system with automated status updates and notifications
- •Integrated with existing tools (Slack, Jira) enabling seamless workflow for cross-functional teams
Time Series Analysis & QC
ML-powered time series analysis and quality control system
- •Developed ML-powered QC system with real-time anomaly detection, reducing data issues by 70%
- •Built specialized visualization engine for time series data, accelerating pattern identification by 55%
- •Implemented automated alert system with severity-based routing, reducing response time by 80%
- •Created quantitative QC scoring system enabling data-driven quality assessment and trend analysis
GenAI Query Canvas
Natural language to visualization pipeline using GenAI
- •Built GenAI-powered query interface reducing data analysis time by 75% through natural language processing
- •Developed automated visualization pipeline (Plotly & D3) with dynamic chart selection and customization
- •Implemented feedback-based learning system improving query accuracy by 40% over time
- •Created end-to-end pipeline from natural language to interactive dashboards, enabling self-service analytics
Machine Learning Developer
NASA Ames Research Center
Mountain View, California, United States
December 2023 - March 2024
ResearchTechnical Stack
PythonPyprojFoliumNext.jsPlotlyDockerAWSNumPyPandasKubernetesGeospatial AnalysisData VisualizationMicroservices
Key Achievements
Performed spatial analysis of flight path adherence to ATC procedures [Python, Pyproj, GeoPandas]Built interactive visualization dashboards for airspace data analysis [Folium, Leaflet.js, Plotly]Identified navigational patterns enabling aviation fuel cost optimization [NumPy, Pandas]Containerized legacy flight analysis system, reducing setup time from days to minutes [Docker]
Flight Path Optimization System
ML-powered flight path analysis and adherence checks using Python, Pyproj, Xmitree and Folium
- •Engineered coordinate transformation pipeline converting flight paths from lat/long to cartesian coordinates using Pyproj, enabling web-based visualization on Folium maps
- •Containerized legacy flight analysis system onto Folium map containers, reducing setup time from days to minutes for new researchers
- •Developed and deployed time series viewing component for flight paths, enabling real-time analysis of flight path adherence
- •Automated data validation pipeline over 1000+ flight paths, reducing manual verification time from 4 hours to 15 minutes per analysis
Graduate Research Assistant
San Jose State University
San Jose, California, United States
July 2023 - February 2024
Research & DevelopmentTechnical Stack
ReactNode.jsBERTGPT-2YOLO v8PythonOpenCVTypeScriptTailwind CSSWix CMSStrapiREST APIsMachine Learning
Key Achievements
Spatial Analytics: ML models on spatial datasets; EDA with R (dplyr, ggplot2), hypothesis testing [Python, R]Motion Planning: Sensor fusion (camera + LiDAR) using KITTI; YOLO v8 object detection [Python, OpenCV]Built multilingual NLP translation system (1-to-25 languages) with LangChain and HuggingFace [NLP, LangChain]Mental health support platform (BERT/GPT-2) for hurricane relief in Hawaii [React, Node.js]
Sensor Fusion & Motion Planning
Advanced sensor fusion and motion planning system
- •Researched and implemented real-time sensor fusion pipeline combining LiDAR and camera data from KITTI dataset
- •Enhanced normalization techniques filtering non-critical data, improving object detection accuracy
- •Recreated and improved YOLO v8 integration with LiDAR data for precise 3D bounding box generation
- •Developed multi-dimensional representation system for motion planning using fused sensor data
OHANA Mental Health Support
Web application for hurricane relief and mental health support
- •Built React-based web application for mental health support and emergency resource location in Hawaii
- •Implemented BERT and GPT-2 models trained on mental health data for emotional support and crisis response
- •Developed geolocation feature to help users find nearest mental health centers and hospitals
- •Created interactive interface for mental health Q&A using transformer models fine-tuned on mental health data
Multilingual TTS Research
Research and evaluation of multilingual TTS systems
- •Researched and evaluated TTS systems supporting 25+ languages using Coqui TTS and Firefox voice synthesis
- •Developed prototype interface for testing and comparing different TTS implementations
- •Created evaluation framework for assessing voice quality and language support across different TTS systems
Senior Software Engineer
Ascendion
Basking Ridge, New Jersey, United States
November 2022 - August 2023
DevelopmentTechnical Stack
AngularTypeScriptAngular 15/16D3.jsHighchartsW3CStorybookAzureDevOpsCI/CDComponent ArchitectureJavaScriptMicroservicesSpring BootMicroservices
Key Achievements
Architected data visualization system processing 10M+ records for enterprise dashboards [Angular, TypeScript]Mentored 20 engineers; conducted 500+ code reviews improving team code quality by 35% [Angular 15/16]Reduced page load times by 85% (23s to 2.5s) through lazy loading and pagination [D3.js, Highcharts]Integrated OpenAI ChatGPT into VS Code extension, boosting developer productivity 25% [OpenAI, Node.js]
Enterprise Angular Migration
Large-scale Angular framework upgrade and modernization
- •Led migration from AngularJS to Angular 15 for legacy systems, improving performance and maintainability
- •Upgraded enterprise applications from Angular 13 to Angular 16, implementing modern features and best practices
- •Created comprehensive migration documentation and automated testing suite ensuring zero regression
- •Implemented modern Angular features (standalone components, signals) reducing bundle size by 35%
Cloud Economics Platform
Real-time cloud resource monitoring and cost optimization platform
- •Developed interactive dashboards using D3.js and Highcharts for real-time VM usage and cost analytics
- •Implemented automated cost optimization recommendations reducing cloud spending by 30%
- •Created REST API integration with AWS & Azure for real-time resource monitoring
- •Built polling system for continuous data updates ensuring accurate cost tracking
AI-Powered Code Quality Tool
VS Code extension integrating ChatGPT for automated code review and quality checks
- •Developed VS Code extension integrating ChatGPT API for automated code review
- •Implemented custom rules engine for code quality assessment and improvement suggestions
- •Created automated documentation generation reducing manual documentation time by 70%
- •Built feedback loop system improving suggestion accuracy over time
Enterprise Component Library
W3C-compliant reusable component library for enterprise applications
- •Architected comprehensive component library following W3C accessibility standards
- •Implemented 40+ reusable components reducing development time by 60%
- •Created detailed documentation and usage examples for 20+ team members
- •Developed automated testing suite ensuring 95% code coverage
Performance Optimization
Enterprise application performance optimization and modernization
- •Reduced page load times from 23s to 2.5s through UI improvements and pagination
- •Implemented efficient data loading strategies reducing API response times by 75%
- •Optimized Angular Material and D3.js implementations for better performance
- •Created performance monitoring dashboard for real-time metrics tracking
Software Engineer
Collabera Inc.
Basking Ridge, New Jersey, United States
December 2021 - December 2022
DevelopmentTechnical Stack
AngularAzureTypeScriptPythonAzure FunctionsAngular MaterialD3.jsW3C StandardsDevOpsSRE
Key Achievements
Built executive dashboards with D3.js/Highcharts, increasing decision-making speed by 75% [D3.js, Highcharts]Engineered reusable Angular component library (50+ components) with WCAG 2.1 compliance [Angular, TypeScript]Developed Azure-integrated ML workflow app, automating 80% of manual data processes [Azure, Python]Awarded Employee of the Quarter for delivering 3 cross-functional projects ahead of schedule [Angular Material]
SRE & DevOps Enterprise Assessment Tool
Automated assessment and monitoring platform
- •Developed Angular Framework-driven assessment tool for Site Reliability & DevOps
- •Implemented automated monitoring and alerting system using Azure Functions
- •Created real-time dashboards for system health and performance metrics
- •Built automated reporting system for operational metrics and KPIs
Enterprise Component Library
W3C-compliant reusable component library
- •Engineered comprehensive Angular component library following W3C accessibility standards
- •Developed 30+ reusable components for forms, tables, and data visualization
- •Created detailed documentation and usage examples for development team
- •Implemented automated testing suite ensuring 95% code coverage
Azure Monitoring Integration
Machine learning workflow integration with Azure
- •Built web application for Azure Backend with integrated machine-learning workflows
- •Implemented real-time data synchronization between frontend and Azure services
- •Developed automated data processing pipeline using Azure Functions
- •Created comprehensive error handling and logging system using Application Insights
Software Developer
LIBS
Cotonou, Littoral, Benin
January 2021 - November 2021
DevelopmentTechnical Stack
AngularTypeScriptAWS EC2REST APIsDatabase DesignSecurityTeam LeadershipInventory ManagementSales Systems
Key Achievements
Client-facing warehouse & sales systemMulti-country inventory managementProduction team leadershipEC2 deployment & security
Warehouse and Sales Management System
Comprehensive inventory, sales, and production management platform
- •Developed real-time inventory tracking system with dynamic stock level monitoring
- •Engineered sales tracking system supporting multiple currencies and languages
- •Led production team of 10+ engineers ensuring timely delivery of products
- •Deployed application on Amazon EC2 with enhanced security across West Africa
Software Engineer
Accenture
Pune, Maharashtra, India
October 2018 - December 2020
DevelopmentTechnical Stack
InformaticaAbinitioUnixMS-VBAPythonNode.jsNoSQLETLData WarehousingBatch ProcessingSQLOracle DBPostmanAWS
Key Achievements
Resolved critical ETL validation issues processing 5M+ daily records with zero data loss [Informatica, Abinitio]Built automation scripts reducing manual reporting effort by 70% (40hrs to 12hrs/week) [Python, Unix]Led production data analysis enabling 3-month early readiness for system switchover [SQL, Oracle DB]
Barclays - Data Migration & ETL
Enterprise data migration and ETL pipeline development for Barclays
- •Developed ETL pipelines using Informatica and Abinitio
- •Optimized batch processing and data workflows
- •Implemented data validation and quality checks
- •Created automated monitoring and reporting system
Frontend Developer
Enthralltech Private Limited
Pune Area, India
May 2018 - October 2018
DevelopmentTechnical Stack
AngularMongoDB.NETAWSNoSQLChatbot DevelopmentREST APIsE-learning PlatformsIAM Systems
Key Achievements
Built e-learning platform features serving 10K+ students using MongoDB and Angular [Angular, MongoDB]Developed chatbot reducing support tickets by 40% through automated query resolution [.NET, NoSQL]
E-Learning Platform & Chatbot
Development of e-learning platform with intelligent chatbot for educational support
- •Built e-learning platform using Angular and MongoDB
- •Developed chatbot using NoSQL for educational Q&A
- •Implemented IAM system for user authentication
- •Integrated AWS services for content delivery
Education
Master of Science, Artificial Intelligence
San Jose State University
San Jose, CA
GPA: N/A
August 2023 - May 2025
Relevant Coursework
- • Advanced Machine Learning
- • Deep Learning
- • Natural Language Processing
- • Computer Vision
- • AI Ethics
- • Knowledge Representation and Reasoning
Achievements
- • Relevant Research Projects (e.g., Geospatial Analysis, Motion Planning)
Bachelor of Engineering - BE, Information Technology
Sinhgad Academy Of Engineering (Pune University)
Pune, Maharashtra, India
GPA: N/A
2014 - 2018
Relevant Coursework
Achievements
Featured Projects
Recent work across AI/ML, data engineering, full-stack development, and developer tooling.