---
title: "Zeeshan Ali - AI Engineer in Bengaluru | Tal"
description: "Meet Zeeshan Ali, AI Engineer @ Visionet Systems Inc. | Building Enterprise Agentic Platform based in Bengaluru. AI stack includes claude code, codex, and cursor. Explore public work links including github."
image: "https://tal.club/meet/zeeshan-ali/share-image"
---

# Zeeshan Ali on Tal

## Headline

AI Engineer @ Visionet Systems Inc\. \| Building Enterprise Agentic Platform

## About

As an AI Engineer at Visionet Systems Inc\., I contribute to the development of AgentVerse, an advanced enterprise\-grade AI agent platform for multi\-agent orchestration\. My work centers on engineering reliable production infrastructure, including agent execution loops, state persistence, and retry mechanisms, which significantly enhance operational efficiency and save engineering hours\. I also built an internal LLM gateway that streamlines multi\-provider integration, optimizes costs, and enforces compliance\. With a B\.Tech in Artificial Intelligence from Presidency University Bangalore, I focus on creating practical, scalable solutions for applied generative AI\. My experience spans building LLM pipelines, designing retrieval\-augmented generation \(RAG\) systems, and implementing observability frameworks to improve system reliability\. My goal is to develop AI tools that solve complex business problems and drive measurable impact\.

## Location

Bengaluru

## AI stack

claude code

codex

cursor

## Work experience

AI Engineer at Visionet Systems Inc\., Feb 2025 \- Present · 1 yr 7 mos, Bengaluru, Karnataka, India, Core contributor to Visionet AgentVerse — a composable, enterprise\-grade AI Agent Platform for multi\-agent orchestration\. • Engineered the production agent harness — the execution loop, tool sandboxing, state persistence, retry/fallback logic, and runaway\-loop guards — shifting reliability from prompt\-tuning to infrastructure and saving 200\+ engineering hours per month\. • Built an internal LLM gateway with multi\-provider routing across OpenAI, Anthropic, and Gemini, semantic response caching, rate limiting, token\-budget enforcement, and PII scrubbing — decoupling product teams from vendor\-specific APIs and cutting LLM spend across services\. • Architected RAG pipelines with hybrid retrieval, re\-ranking, metadata filtering, structured outputs, and layered memory \(short\-term, episodic, long\-term\) over Pinecone and pgvector — turning unstructured enterprise data into governed, agent\-callable knowledge\. • Designed a low\-code agentic workflow builder with human\-in\-the\-loop approval gates, policy\-aware tool routing, schema\-validated outputs, and hallucination guardrails — enabling non\-technical teams to ship autonomous workflows safely\. • Stood up the agent observability stack with OpenTelemetry causal tracing \(LangSmith, Langfuse\), Ragas and LLM\-as\-judge evaluations, golden\-set regression suites, and drift detection \(MLflow, Prometheus, Grafana\) — improving deployment reliability by 40% with full decision provenance for regulated environments\. Stack: Python, LangChain, LangGraph, LlamaIndex, MCP, A2A, AWS Bedrock, Vertex AI, Azure AI Foundry, Pinecone, pgvector, Redis, Docker, Terraform, OpenTelemetry, Langfuse\.

Co\-Founder at Quanta, Mar 2024 \- Present · 2 yrs 6 mos, Bengaluru, Karnataka, India, Co\-founded Quanta to ship applied generative AI products that solve real business problems with measurable impact — not demos\. What I own: • Set the technical direction and architecture for Quanta's GenAI product surface, from prototype to production\. • Designed end\-to\-end LLM pipelines: retrieval, prompting, evaluation, deployment, and cost monitoring\. • Built tiered inference architectures \(rules → classic ML → LLM\) to keep unit economics workable as usage scales\. • Owned the trade\-off conversations with stakeholders: when to fine\-tune vs\. prompt, when to call an LLM vs\. avoid one, and where reliability matters more than capability\. Focus: applied generative AI for productivity workflows — building tools where the model is a component, not the whole product, and where every inference request answers "can this be solved without an LLM?" before calling one\. Stack: Python, LangChain, OpenAI/Anthropic APIs, vector databases, FastAPI, Docker, AWS\.

Co\-Founder at HUANAR, Oct 2022 \- Dec 2023 · 1 yr 3 mos, Bengaluru, Karnataka, India, Co\-founded HUANAR, an EdTech venture blending AI with learning and development to make personalized education accessible\. What I owned: • Set product direction and tech architecture from zero — user research, MVP scoping, build, and iteration with early users\. • Designed and shipped the learning experience: UI/UX, content delivery, and interaction model\. • Integrated AI into the L&D loop: automated content recommendations, learner\-progress signals, and personalization driven by usage data\. • Wore the full early\-stage hat: customer conversations, hiring early collaborators, and prioritising the smallest set of features that demonstrably moved learner outcomes\. Why it matters: HUANAR was where I learned to think in user outcomes, not features\. Every technical decision was filtered through one question — "does this actually help someone learn faster?" That product instinct is what I now bring to every AI system I ship\. Focus: EdTech, AI\-assisted learning, UX design, product\-led iteration\.

Software Developer at Vikasana, Mar 2022 \- Jan 2023 · 11 mos, Bengaluru, Karnataka, India, Full\-stack development internship building production\-grade web applications and learning the discipline of shipping code that real users depend on\. What I owned: • Built and shipped end\-to\-end web features — frontend \(HTML/CSS/JavaScript\) and backend \(REST APIs, database integration\)\. • Translated product requirements into technical specs and working code under real deadlines\. • Debugged across the stack: from CSS layout issues to backend query performance\. • Collaborated with designers, PMs, and senior engineers in a structured agile workflow — stand\-ups, code reviews, sprint planning\. Key takeaway: this internship gave me the engineering fundamentals — version control, testing, deployment, code review — that I now apply to every AI system I build\. Production discipline starts here, not in ML libraries\. Stack: HTML, CSS, JavaScript, Python, REST APIs, Git, SQL\.

## Education

B\.tech, Artificial Intelligence, Presidency University Bangalore

## Public links

- github: https://github\.com/zeeshanunique

[View the original profile](https://tal.club/meet/zeeshan-ali)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/zeeshan-ali#person",
      "name": "Zeeshan Ali",
      "url": "https://tal.club/meet/zeeshan-ali",
      "image": "https://tal.club/meet/zeeshan-ali/profile-photo.jpg",
      "description": "Meet Zeeshan Ali, AI Engineer @ Visionet Systems Inc. | Building Enterprise Agentic Platform based in Bengaluru. AI stack includes claude code, codex, and cursor. Explore public work links including github.",
      "jobTitle": "AI Engineer",
      "worksFor": {
        "@type": "Organization",
        "name": "Visionet Systems Inc."
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "AI Engineer"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bengaluru"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Presidency University Bangalore"
        }
      ],
      "knowsAbout": [
        "claude code",
        "codex",
        "cursor"
      ],
      "sameAs": [
        "https://github.com/zeeshanunique"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/zeeshan-ali#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/zeeshan-ali#webpage",
      "url": "https://tal.club/meet/zeeshan-ali",
      "name": "Zeeshan Ali on Tal",
      "description": "Meet Zeeshan Ali, AI Engineer @ Visionet Systems Inc. | Building Enterprise Agentic Platform based in Bengaluru. AI stack includes claude code, codex, and cursor. Explore public work links including github.",
      "dateModified": "2026-08-29T05:34:44.594Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/zeeshan-ali#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/zeeshan-ali/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/zeeshan-ali/profile-photo.jpg",
        "caption": "Zeeshan Ali profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```