Gulab Ahmad
Software Engineer | Backend & AI Application Engineer
Hire this person
Send a job offer directly to this candidate
About
Backend-focused Full Stack Engineer specializing in enterprise SaaS platforms, multi-tenant architecture, and AI-powered applications, with delivery experience across healthcare, logistics, customs clearance, automotive commerce, financial systems, and AI SaaS. Designed and developed 550+ REST APIs and 150+ relational database tables across four multi-tenant SaaS systems as the sole engineer on three of them. Core strengths in Python/FastAPI backend architecture, relational schema design, and secure authentication (OAuth, JWT, RBAC) with tenant-aware data isolation, complemented by hands-on AI engineering across OpenAI, Gemini, RAG, and VAPI voice agents.
Comfortable-owning delivery end-to-end from schema design through Docker/AWS EC2 deployment and GitLab CI/CD.
Experience
Software Engineer - Progmize
(2025-11)
Product software company building enterprise SaaS platforms for logistics, automotive, and auction businesses.
Software Engineer - Clearance Management System (CMS) - Progmize
(2025-11)
Multi-tenant UAE port import/export clearance platform.
Problem: UAE port clearance ran across disconnected tools documents, driver assignments, and payments tracked separately causing delays, demurrage exposure, and no shared visibility for operators, customers, and vendors.
Architecture: Architected a modular FastAPI backend split by domain (clearance, transport, finance, documents, notifications, access control) over a normalized 40–50 table PostgreSQL schema; chose shared-schema multi-tenancy with tenant_id isolation and tenant-aware queries to serve many tenants from one database while keeping their data separated, and layered RBAC via JWT/OAuth for five roles. Implementation & business logic: Developed 180+ REST APIs modeling the full container lifecycle (arrival → clearance → delivery → return → settlement → refund) and a financial engine demurrage/toll calculation, refundable customs deposits, delivery-order generation, refunds wired to email, SMS, and payment-gateway services for automated notifications and collection.
Outcome: Containerized with Docker and deployed on AWS EC2 behind Nginx via GitLab CI/CD, replacing a fragmented process with a single source of truth acros
Software Engineer - AutoPartsX - Progmize
(2025-11)
Automotive-parts e-commerce SaaS. Problem & architecture: A parts business needed to run international sales alongside inventory, procurement, and accounting; designed a modular backend around a VIN-based part-compatibility engine the core design choice so catalog, stock, purchasing, and ledger stay in sync instead of behaving as isolated features. Implementation & business logic: Developed 130+ FastAPI REST APIs and a 38-table relational schema spanning inventory, purchase orders, GRN, suppliers, customers, branch transfers, ledger, and finance, orchestrating a procurement → inventory → finance → invoice-generation workflow so a purchase flows through to stock, accounts, and invoicing without manual reconciliation.
Outcome: Delivered end-to-end as sole engineer from schema design through Docker/AWS EC2 deployment a production-ready SaaS unifying storefront and back-office in one system.
Software Engineer - Smart Yard Auction OS - Progmize
(2025-11)
Back-office OS for UAE auction yards (MVP). Problem & build: UAE auction yards ran on paper logs and spreadsheets, making inventory error-prone and VAT accounting hard to keep compliant; delivered a modular full-stack MVP 90+ FastAPI REST APIs over a 30-table schema covering auction inventory, workflow, and accounting owed from database design to release.
Business logic: Implemented a UAE Profit Margin Scheme (PMS) VAT engine with automated commission and invoice generation, encoding FTA tax rules directly into the workflow so compliance is enforced by the system rather than manual bookkeeping.
Outcome: Shipped as a containerized deployment on AWS EC2, giving yards structured digital inventory and an automated, compliant financial engine in place of manual record-keeping.
AI Engineer - Visibility Bots
(2025-11 - 2026-04)
Independently designed and built three AI applications spanning support automation, research generation, and content tooling.
AI Engineer - Research Paper Agent - Visibility Bots
(2025-11 - 2026-04)
Problem & workflow: To produce trustworthy research output, built a tool that turns a query into structured, source-cited essays via a RAG workflow retrieving from an embeddings-backed vector database before generation, so answers stay grounded in real sources rather than hallucinating. Implementation: Built the OpenAI SDK pipeline (embedding → retrieval → prompt orchestration) behind a React frontend, delivering interactive, citation-backed research.
AI Engineer - AI Contact Support App - Visibility Bots
(2025-11 - 2026-04)
Problem & workflow: Offloaded first-line customer support with a Gemini-powered agent that resolves FAQs, refund, return, and order-tracking queries, using context-aware prompting to keep answers scoped to each request. Implementation: Structured the conversational workflow and frontend around common support intents, producing predictable, on-topic responses that reduce repetitive manual handling.
AI Engineer - YouTube Script Agent - Visibility Bots
(2025-11 - 2026-04)
Problem & workflow: Cut manual scripting effort with an LLM workflow that analyzes a topic and generates structured, ready-to-use scripts. Implementation: Designed the topic-analysis and content-planning steps and deployed the app for live use.
Full Stack Engineer - FISTA Solutions
(2024-01 - 2025-11)
20-person product company delivering AI-powered SaaS platforms for healthcare and enterprise training.
Full Stack Engineer - MetaCura (Healthcare Management SaaS) - FISTA Solutions
(2024-01 - 2025-11)
Problem: A multi-tenant healthcare platform needed to centralize fragmented records across hospitals, pharmacies, and patients and make dense clinical reports understandable to non-technical users. Architecture & ownership: Owned authentication and multi-tenancy implemented shared-schema tenant_id isolation with tenant-scoped queries and RBAC (OAuth/JWT) to keep each tenant's medical data separated in a shared database plus the patient and dashboard APIs and database performance (query optimization, indexing) within a 150+ API, 47-table system.
AI layer: Built an OpenAI-powered RAG chatbot grounded in patient records and a VAPI real-time voice assistant that summarizes clinical reports in natural language; deployed via GitLab CI/CD on Docker, AWS EC2, and Nginx.
Frontend-focused Full Stack Engineer - Oliver AI Live Coaching - FISTA Solutions
(2024-01 - 2025-11)
AI voice-coaching SaaS. Built the multi-tenant SaaS frontend and frontend architecture for a platform that trains and evaluates employees through real-time VAPI voice-coaching conversations, integrating the UI with the AI voice layer. Owned API integration, authentication, user management, and product-management workflows, connecting frontend state to backend services and the voice pipeline.
Education
B.S. - Computer Science - Government College University, Faisalabad (2021 - 2025)

