Hi, I'm Deepak Singh.

Expanded from iOS into backend and full-stack AI — building and running AI products in Python, TypeScript and PostgreSQL, with Claude Code and OpenAI Codex as coding agents.

• Insy.ai — agentic RAG HR assistant (in production). Built the retrieval pipeline (document parsing, chunking, OpenAI embeddings, pgvector search) and agent tool calling over live ERP data inside the company's Odoo 14 HR ERP, shipped as an isolated module without touching ~64 existing production modules.
• VaSi.ai (Learnee) — analytics platform (in production). Built a Firebase Functions backend of scheduled agents pulling from Cloud Logging, BigQuery/GA4, Elasticsearch and internal APIs into PostgreSQL, with rule-based monitoring, Claude-written explanations with fallbacks, and JWT role-based access.
• Own the full lifecycle — architecture, specs and implementation plans, code review of agent-written code, deployment and production support.

Lead the iOS practice across two flagship products — Sully.ai (AI clinical assistant) and WiSA (wireless audio hardware) — owning team direction, technical architecture, and stakeholder partnership end-to-end.

• Lead the Sully.ai iOS team of 3 — own roadmap prioritization, release planning, and stakeholder alignment. Decide what ships each cycle based on customer impact and engineering risk; close the loop with product, design, backend, and clinical operations.
• Hired to fix Sully.ai's critical audio-corruption issue — shipped the fix. Rebuilt the audio capture and PCM streaming layer (AVAudioEngine + WebSockets) to deliver clean recordings into the OpenAI realtime transcription, translation, and clinical-note pipeline.
• Built a reusable Swift BLE framework for WiSA (IC) — abstracts ESP32 microcontroller and WiSA transmitter communication into a clean API. Powers the iOS app that controls WiSA-ready hardware across 40+ audio brands including Bang & Olufsen, Harman Kardon, Klipsch, and LG.
• Client-facing technical lead — run discovery and architecture conversations directly with WiSA stakeholders. Translate audio-hardware constraints into mobile workflows and push back on scope when timelines aren't realistic.
• Cut release prep from 1+ hour of manual work to a single-click deploy by automating signing, TestFlight, and App Store delivery via Fastlane + GitHub Actions.
• Established production observability — introduced LogRocket + Firebase structured logging, audio-pipeline instrumentation, and a triage rotation that surfaces issues before users report them.
• Standardized AI-augmented workflows (Cursor, Claude, MCP) across the iOS team — accelerating code review, debugging, and documentation.

Took two products from zero to App Store while growing from individual contributor to team lead — culminating in managing a cross-platform mobile team and owning client relationships.

• FactorLab / SmartTagIt — built and led mobile (iOS + Android). Took SmartTagIt from zero to App Store as the iOS owner, then expanded into managing a 3-person cross-platform mobile team. Ran planning, code review, hiring input, and direct client communication.
• IoT data pipeline at scale — engineered offline-first iOS workflows capturing millions of sensor data points per day with background uploads to cloud streaming infrastructure. Chose Realm over Core Data after benchmarking — write-heavy sensor data made the difference.
• Clove Dental — shipped Patient & Doctor apps from scratch. Owned the full lifecycle: architecture, REST + socket-based realtime communication, Core Data persistence, secure cloud file storage, App Store submission, and post-launch iteration. Apps power Asia's largest dental chain.
• Technical voice in client meetings — defined scope, estimated, surfaced risk early, and proposed alternatives when requirements clashed with mobile platform realities.
• Built patterns that outlived the project — reusable networking, caching, and crash-debugging conventions adopted as the default starting point for new iOS work.

Contributed to Swagbucks (5M+ users) and reusable framework components spanning video streaming, surveys, feedback, advertising monetization, and partner integrations.

• Developed and maintained multiple production iOS applications using Swift, UIKit, Objective-C, VIPER, REST APIs, and iOS SDK frameworks.
• Built and optimized video streaming workflows for high-volume media consumption — improving playback experience, loading performance, and stability.
• Designed feedback and survey features using VIPER, improving user interaction and structured data collection.
• Developed and maintained an advertising iOS framework distributed to B2B partners for monetization integrations.
• Built reusable UI components, networking layers, and shared utilities to support faster feature development across multiple apps.