Dubai, UAE · Senior AI Engineer & Team Lead

Salman
Zafar

I take LLM-powered systems from prototype to production —agentic workflows, RAG pipelines, and multi-agent orchestration— and lead the teams that ship them.

// building production agentic systems
Get in touch →GitHubLinkedIn
10+ yrs shipping
500+
users on enterprise platform
90x
faster critical responses
35%
lower LLM inference cost
10+
years in software engineering
01 / About

Engineering AI
that ships.

Senior AI engineer and technical team lead specializing in production-grade applied AI and agentic systems.

I design, build, and deploy LLM-powered solutions end to end — agentic workflows, RAG and GraphRAG pipelines, multi-agent orchestration, and customer-facing AI delivery. I pair deep hands-on engineering with team leadership, owning technical architecture and standards while translating business problems into systems that hold up under real-world load, cost, and safety constraints.

I'm also the founder of promptev.ai, a no-code agentic AI orchestration platform.

Production AIAgentic SystemsTeam LeadershipMENA Region
02 / Capabilities

The stack I build with

Applied AI & Agentic Systems
Multi-agent orchestrationTool / function callingReActPlan-and-ExecuteLLM integrationContext engineeringPrompt engineering
Models & Providers
OpenAIAnthropic / ClaudeGoogleMistralOpen-source modelsDynamic routing
Agentic Frameworks
LangChainLangGraphAutoGenCrewAILlamaIndexSemantic KernelOpenAI Agents SDKAzure AI Foundry
RAG & Retrieval
GraphRAGMultimodal RAGAgentic RAGEmbeddingsPineconeWeaviateMilvuspgvectorHybrid searchRe-ranking
MCP & Multimodal
MCP servers & toolsMultilingual Context Engine / PipelineOCR & extractionAzure Doc IntelligenceClaude visionDocument · image · audio
AI Safety & Evaluation
GuardrailsContent filteringPrompt-injection defenseHuman-in-the-loopAccuracy · safety · latency evals
Languages & Frameworks
Python / FastAPINode.js / ExpressTypeScriptReact.jsNext.js
Data & Infrastructure
PostgreSQLRedisAWS · Azure · GCPDockerCI/CDTerraformPulumiObservability
Leadership
Team leadership & mentoringTechnical architectureStandards & deliveryStakeholder management
03 / Experience

Where I've worked

Publicis GroupeCURRENT
2022 — Present · Dubai, UAE
Senior AI Engineer / Engineering Team Lead
  • Lead design, build, and deployment of LLM-powered agentic systems across an enterprise platform serving 500+ users.
  • Drive AI/GenAI adoption org-wide through Internal Digital Transformation (IDT).
  • Built RAG/GraphRAG pipelines, evaluation frameworks, and a multimodal Context Engine.
  • Lead a team of engineers; own technical standards and delivery.
promptev.aiCURRENT
2025 — Present · Dubai, UAE
Founder & Lead Engineer
  • Built a no-code agentic AI orchestration platform end-to-end: GUI agent builder + reusable Context Engine (75+ languages) with multimodal RAG ingestion.
  • Security-first BYOK architecture; connects agents to enterprise tools via user-configured MCP servers.
  • Published SDKs on PyPI & npm, a CLI, and a VS Code extension adopted by developer communities.
TeqXS
2020 — 2022 · Lahore, Pakistan
Senior Software Engineer
  • Optimized AWS infrastructure to 99.9% uptime and cut deployment time by 50%.
  • Designed relational and NoSQL data models; partnered with mobile, marketing, and front-end teams.
Vfairs
2021 — 2022 · Lahore, Pakistan
Team Lead / Senior Full-Stack Developer
  • Led and mentored a cross-functional team of 6 on a large-scale virtual-events platform.
  • Architected a maintainable platform at 99.9% uptime with modular REST APIs.
Innowi Inc.
2017 — 2022 · Lahore, Pakistan
Software Engineer
  • Built full-stack web portals and IoT-enabled systems with RESTful APIs for real-time device communication at 99.9% uptime.
  • Employee of the Year 2018.
04 / FlagshipFounder

promptev.ai

A no-code agentic AI orchestration platform — build, deploy, and manage AI agents without writing code, on a proprietary DRAFT framework (Definition, Role, Action, Format, Task/Tone).

Visit promptev.ai ↗
DRAFT framework
Definition · Role · Action · Format · Task/Tone — a structured method for building reliable agents.
BYOK multi-model
Bring your own keys across providers; route to the right model per task.
Context Packs & RBAC
Reusable context bundles with access control down to route, button, and tab level.
Multiple deployment modes
Web widget, private link, embed, API, SDK, Slack/Teams, and scheduled autonomous agents.
Human-in-the-loop gates
Inline and email-based approval per tool for safe autonomous execution.
MCP client & tooling
Connects to user-provided MCP servers. Published SDKs on PyPI & npm, a CLI, and a VS Code extension.
05 / Impact

Measurable outcomes

Enterprise AI platform
Architected and delivered a platform serving 500+ users across 7 modules in the MENA region.
35% lower inference cost
Cut LLM cost through model selection, dynamic routing, and caching.
Up to 90x faster
Optimized critical API and process response times by up to 90x.
2.5x less overhead
Reduced process overhead by 2.5x through automation and re-architecture.
Multimodal Context Engine
Ingests and extracts structured data from documents, images, and audio (Azure Document Intelligence + Claude OCR).
Enterprise-grade security
RBAC, PII redaction, encryption, and audit logging for regulated, sensitive-data environments.
06 / Writing

Recent writing

All posts →
07 / Contact

Let's build something
intelligent.

Open to senior applied-AI roles, technical leadership, and agentic-systems collaborations.

salmanzafar949@gmail.comGitHubLinkedIn
BSCS — University of South Asia, Lahore (2012–2016)Languages: English · Urdu