Baldeep Singh Kwatra
AI-first engineering leader & product builder
I build AI products and help engineering teams work with AI. I stay close to the code, the architecture and the decisions that turn an idea into a running business.
Founder & engineering leader
2012–2015 / DRIKU
DRIKU was a chauffeur-on-demand service in Delhi, available through mobile apps and the web.
I founded the company and built its technology while studying computer science.
Read about DRIKU ↘2020–2025 / FarEye
I replaced tightly coupled integrations with reusable connectors that teams could deploy independently.
My role expanded from architecture and implementation to leading Integration, DevOps, SRE and Platform Engineering.
Explore the FarEye work ↘2025–Present / Tractable
I built the initial AI vehicle-inspection platform, including the application, inference integration, infrastructure and deployment.
I used AI coding agents throughout development and reviewed the architecture, implementation and production behavior myself.
Read the LumaScanner case study ↘DRIKU / Jan 2012–May 2015
customers served in Delhi
Founder & CTO · alongside my degreeI founded and developed a chauffeur-on-demand platform across Android, iOS and the web.
I designed the architecture for reuse, so it could support other on-demand services with minimal customization.
01 / Perspective
I treat AI as part of how engineering works: in the products we build, the code we write and the way we review decisions.
I introduce AI-assisted practices through hands-on implementation, shared review standards and mentoring. I want engineers to understand the systems they ship and use AI with informed judgment.
I remain accountable for architecture, reliability, cost and the people doing the work. AI expands what a team can build; engineering experience guides what we choose to build and how we operate it.
Selected achievements
02 / Selected work
An AI vehicle-inspection platform, built end to end.
I personally designed and implemented the initial platform: frontend, backend, AI inference integration, cloud infrastructure and production deployment. The first production customers launched in four weeks.
I own the technical roadmap and continue to implement major capabilities, including notifications and reporting. I also work on customer integrations, including dealership management systems.
Inspection pipeline
Processing overview
A TypeScript monorepo connects the application stack. Temporal provides durable workflow orchestration; Kubernetes and Helm support deployment. PostgreSQL, Redshift and Redis serve the platform’s data needs.
I built proprietary-model and Gemini-backed inference paths, with labeled-data evaluation and precision/recall comparisons. I select models for different tasks based on quality, latency and cost, and review model providers for risk and suitability.
I operate the platform on AWS, with GitHub Actions for delivery and Grafana, Loki and Tempo for observability. My responsibilities include production incident resolution, engineering security decisions and enterprise security and privacy reviews.
After the initial implementation, I onboarded and mentored three engineers to extend the platform using the established architecture, reviewed plans and release process.
I conceived, architected and built FarEye’s first company-wide Kubernetes-based product: an integration platform that replaced tightly coupled integrations with reusable connectors.
The architectural change gave integrations their own deployment boundaries and teams more control over delivery.
After: reusable connectors deploy and scale independently. A change to one connector does not require a shared release.
See which connectors need a release when Connector B changes.
Simplified architecture
Connectors reuse common integration logic instead of rebuilding it for each onboarding.
Independent deployments and scaling reduce coupling and improve operational reliability.
Clear component ownership lets engineers make changes without coordinating every integration’s release.
FarEye / Engineering scope
Alongside the integration product, I took responsibility for the infrastructure and engineering systems used across FarEye. The work covered API integrations, delivery tooling, multicloud operations and production reliability.
I designed services for label generation, shipment tracking, webhook processing and file transfers. Reusable connectors combined API calls, SFTP transfers and S3 uploads into configurable workflows, supporting both polling and webhook-based integrations.
I selected Spring Boot, GraalVM or Go according to the workload, with PostgreSQL, Redis and Elasticsearch serving different storage and retrieval needs. The integration platform ran on AWS and GCP Kubernetes.
I built an internal developer platform for self-service releases, environment configuration and deployments. An umbrella Helm-chart model packaged dependent components together, with tagging and GitHub Actions automation standardizing releases.
This work extended beyond individual services to the tools and release processes engineers used across the organization.
I managed Kubernetes operations across AWS, GCP, Azure and Oracle Cloud, including major and minor cluster upgrades, maintenance and security patching. I introduced Terraform infrastructure as code and led high-level cloud governance improvements.
To handle rapid demand spikes, I improved node provisioning and pod scheduling, moving from EC2 autoscaling toward EC2 Fleet and Karpenter for faster capacity provisioning.
I established SRE capability for round-the-clock monitoring and introduced improvements to observability, distributed tracing and alerting. The work included Loki and Grafana, incident-management tooling and production diagnosis.
I led infrastructure cost optimization through changes to compute, capacity planning and purchasing. My responsibilities included budgets, vendor decisions and investment proposals.
Reported annual AWS savings at FarEye.
Reduction in company-wide Tractable infrastructure expenditure.
Graviton, Spot Instances, Karpenter and workload rightsizing helped improve resource use. Resource cleanup, Savings Plans and Reserved Instances complemented the technical changes.
I own global engineering infrastructure spending and participate directly in financial decisions, vendor selection, infrastructure contracts, purchasing approvals and engineering resource allocation.
I owned company-wide AWS infrastructure expenditure, vendor approvals, infrastructure contracts and purchasing decisions. I presented budgets and technical tradeoffs to the CTO and principal architect, and helped finance colleagues understand infrastructure costs.
I established Tractable’s India Development Centre, set interview standards, hired and onboarded early engineers, and helped set up its operations. The hiring process reduced time-to-hire by a reported 40%.
At FarEye, I led more than 40 engineers across Integration Engineering, DevOps, SRE and Platform Engineering, with four engineering managers and a principal-level engineer reporting directly to me.
I mentored engineers into technical leadership and management through coaching, training and architecture discussions. At FarEye, I developed Platform Engineering from the DevOps function and coached engineers as they took on team leadership responsibilities.
I worked with the CTO and product leaders on annual technology plans and quarterly roadmaps. I also supported enterprise sales through solution architecture and RFP responses.
I make decisions about team structure, engineering investment and delivery priorities, and explain the tradeoffs to the people responsible for execution.
03 / AI-first engineering in practice
At LumaScanner, I developed a workflow that combines AI-assisted implementation with human-reviewed plans, code reviews and automated quality checks.
Architecture & scope
Architecture and implementation plans are reviewed by a human before coding begins. Approved plans live in the repository as persistent context for engineers and agents.
Human checkpoint Problem, architecture and scope approved.
Implementation
I use multiple models and coding agents, including Claude, Cursor, Copilot and Codex. Implementation follows the approved plan; I remain hands-on with architecture and major changes.
Human checkpoint Review implementation choices and system fit.
Code review & CI
A custom local pre-PR review applies standards drawn from my own code reviews. I use a different model for an independent perspective, followed by peer review, human approval and CI quality gates.
Human checkpoint Resolve findings and approve the change.
Deployment & operations
GitHub Actions, container builds and Helm automate Kubernetes delivery. Self-hosted runners, merge queues and dependency automation support the release process. Logs, traces and production feedback guide investigation and subsequent changes.
Human checkpoint Review production behavior and resolve issues.
04 / Technical depth
TypeScript, JavaScript, Java, Node.js, Spring Boot, Go, GraalVM, REST APIs, microservices and event-driven architecture. Frontend experience includes Vue.js, Angular and React; earlier application work includes Hibernate and multi-tenant SaaS.
AWS, GCP, Azure, Oracle Cloud, Kubernetes, Docker, Helm, Terraform, Karpenter, GitOps and CI/CD.
PostgreSQL, Redshift, MySQL, MongoDB, Redis, Kafka, RabbitMQ, Elasticsearch, Temporal, Prometheus, Grafana, Loki and Tempo.
Computer vision and vision-language inference, labeled-data evaluation, model quality and cost tradeoffs, human-reviewed AI-assisted development.
Director of Engineering
AI product development, the India Development Centre and global engineering infrastructure budgets.
Senior Engineering Manager · Apr 2024–Jun 2025
Engineering Manager · Jun 2022–Mar 2024
Technical Lead · Apr 2020–Jun 2022
Senior System Analyst
Built Worbix marketing automation and analytics: social publishing, email scheduling and integrations using AWS serverless services, Vue.js and Spring Boot. Work spanned Lambda, DynamoDB, SQS, SNS, SES, S3 and API Gateway, alongside integrations with SendGrid, Redtail and Stripe.
Recognition: Wings to Fly Award.
Senior Software Engineer
Developed multi-tenant SaaS and telecom BI / ETL solutions with Java, Spring, Hibernate, MySQL and Jasper tools.
Recognition: Client Accreditation.
Project Engineer
Automated defense and telecom workflows using Java and IBM Rational Functional Tester.
Recognition: Topper’s Award.
Full-stack development at Amity University on shuttle, pharmacy and assessment platforms; claims workflow work at Religare in 2014. Earlier trainee software-development experience at Progia.
06 / Recognition & writing
My experience moving from Drupal to Strapi and choosing a headless CMS.
Read on Strapi ↗Logistics category, India DevOps Show. Recognition for the FarEye team I led.
View recognition ↗Profit Powerhouse, Challenger, Superpower of Innovation & Creativity, and Team Excellence awards.
View award ↗Education
Amity University, Noida · 2015 · CGPA 8.8/10
AI Club President, 2013–2015.
Young Achiever & Best Technology Creation recognition.
Hillwoods Academy, Delhi: Class 12 in 2011 and Class 10 in 2009. Recognition included a Certificate of Merit in Information Technology, a scholarship, a Delhi Sanskrit Academy award and a science-exhibition prize.
College representative for IIT Bombay’s Entrepreneurship Cell, campus ambassador for CETPA InfoTech, and a delegate at the Microsoft Intelligent Cloud event in Delhi. I enjoy public speaking, technical writing and training.
I speak English, Hindi and Punjabi, with basic French and Sanskrit. Away from work, I enjoy badminton and walking by the sea.
07 / In their words
“Baldeep has been a highly dependable, hands-on leader. He is one of the few Engineers who can relate very well with Business, Customer and Priorities. He is never shy of trying new technology, implementing state-of-art architectures while keeping his team motivated.”
“It is seldom to find people who are as talented as Baldeep. At every opportunity I got to work with Baldeep since 2014, he stood apart every time with his innate ability to assess the requirements and deliver a product that not only enhanced the productivity of the company but reduced the operational cost drastically.”
“He is an asset to any company in the world to fight for.”
“I have had the privilege of working with Baldeep and can confidently say he represents that rare breed of professional who excels across the entire technology spectrum. His unique ability to operate seamlessly from hands-on coding and system design to enterprise architecture and executive engagement sets him apart.”
Read the rest of Balaji’s recommendation
“What truly distinguishes Baldeep is his exceptional talent for transforming challenges into strategic opportunities. He doesn't just manage situations; he elevates them, consistently delivering outcomes that exceed expectations. His work in transforming high-scale integration systems and platform engineering is particularly noteworthy, where he has masterfully implemented automated DevOps practices while maintaining rigorous security standards and optimizing AWS infrastructure costs.”
“Baldeep combines deep technical expertise with outstanding professional discipline. Throughout our collaboration, I've been consistently impressed by his unwavering commitment to deadlines and deliverables. He is that rare professional who can engage with technical teams, architect enterprise solutions, and communicate effectively with leadership, all while driving meaningful business results. I would welcome the opportunity to work with Baldeep again and highly recommend him for any organization seeking transformative technical leadership.”
“Baldeep is a thorough professional to work with and has good technical exposure. He is a sharp, self-motivated and self-managed professional who always strives to follow best practices and optimized approach. I feel he has good leadership potential.”
“Would always like to have a team member like him in my team.”
“Baldeep is skilled in cloud computing (particularly in AWS using EC2 and serverless) and development using Java and related frameworks and libraries. I worked with him in creation and integration of REST based web services for one of the products.”
“It's been pleasure working with such a sincere and dedicated team mate.”
Roles as originally attributed.
08 / Start a conversation
For engineering leadership opportunities or conversations about AI products and platforms, send me a message.