--- title: "Bhavarth Bhangdia - Machine Learning Engineer in India | Tal" description: "Meet Bhavarth Bhangdia, Applied AI & Machine Learning Engineer | Generative AI | Agentic Workflows | Scaling Image & Video Models (Diffusion/Temporal) to 6M+ Users | IIIT Allahabad | AI Engineers based in India. AI stack includes antigravity, cursor, and opencode." image: "https://tal.club/meet/bhavarth-bhangdia/share-image" --- # Bhavarth Bhangdia on Tal ## Headline Applied AI & Machine Learning Engineer \| Generative AI \| Agentic Workflows \| Scaling Image & Video Models \(Diffusion/Temporal\) to 6M\+ Users \| IIIT Allahabad \| AI Engineers ## About I am a Machine Learning Engineer focused on building production\-grade AI systems that turn frontier model capabilities into reliable, scalable products\. My work sits at the intersection of Generative AI, agentic workflows, RAG systems, inference optimization, and ML infrastructure, with a strong focus on measurable business impact\. Over the past 2\+ years, I have worked on AI products from early discovery to production launch, partnering with product and engineering teams to solve ambiguous problems and ship systems that are fast, dependable, and useful in real\-world environments\. Career Highlights: • Optimized distributed Generative AI inference pipelines at Pixlr, reducing latency by 76% from 33s to 7\.9s while supporting millions of users across image and creative AI workflows\. • Built SkillMesh, a LangGraph\-based multi\-agent orchestration system that improved task success rate by 42% and reduced manual intervention by 47% through structured planning, tool use, memory, and evaluation loops\. • Developed enterprise RAG and evaluation pipelines across 100K\+ documents, improving retrieval precision by 28%, reducing hallucination rate by 35%, and enabling systematic measurement of faithfulness, answer quality, and token\-level cost\. I enjoy building AI systems that are not just impressive in demos, but reliable under production constraints\. My core interests include agentic AI, enterprise RAG, LLM infrastructure, AI evaluation, multimodal AI, inference optimization, and product\-focused machine learning\. Core Competencies: Machine Learning, Generative AI, Applied AI, Agentic AI, LLMs, RAG, Multi\-Agent Systems, LLM Evaluation, Prompt Engineering, PyTorch, Python, LangGraph, LangChain, Transformers, Diffusers, Stable Diffusion, SDXL, ControlNet, Vertex AI, GCP, AWS, Ray Serve, FastAPI, Docker, Redis, Vector Databases, MCP, ML Infrastructure, Inference Optimization, Production ML Achievements: Competitive programmer with a Codeforces rating of 1900\+ \(Candidate Master\), reflecting strong problem\-solving, algorithmic thinking, and proficiency in data structures and algorithms\. ## Location India ## AI stack antigravity cursor opencode ## Work experience Machine Learning Engineer at Pixlr Online AI Photo Editor, Aug 2024 \- Jun 2026 · 1 yr 11 mos, Singapore, \- Architect AI infrastructure and production\-grade generative AI systems for Pixlr’s flagship products, leading the development of agentic workflows, enterprise AI features, and scalable inference pipelines\. \- Designed and shipped a skill\-based multi\-agent orchestration system using LangGraph, LangChain, and MCP\-style tool interfaces, with dynamic task routing, hierarchical delegation, ReAct reasoning, shared memory, and self\-reflection, reducing manual intervention by 47% and improving task completion quality\. \- Built observability, tracing, and evaluation frameworks for complex agentic workflows, including trace inspection, failure analysis, and golden dataset regression testing to improve system reliability and support continuous iteration\. \- Optimized prompt pipelines and agent execution through prompt batching, context reuse, adaptive routing, and resilient retry strategies, cutting median latency by 33% and cost per successful request by 25% while maintaining output quality under production traffic\. \- Led AI infrastructure for Pixlr’s flagship AI Designs tool, scaling Stable Diffusion and SDXL\-based generation pipelines, including LCM, LoRA, and ControlNet, to support interactive experiences for millions of users under strict latency constraints\. Artificial Intelligence Engineer at Scale AI, Dec 2023 \- Mar 2024 · 4 mos, Designed large\-scale evaluation pipelines for code generation models spanning 75K samples, formalizing correctness criteria and acceptance thresholds to handle ambiguous and partially\- orrect outputs at production scale\. Accelerated evaluation throughput by replacing naive parallelization with cach \-aware batching strategies, cutting p99 evaluation runtime from 74 minutes to 36 minutes while preserving result consistency and reviewer trust\. Data Science Intern at Stealth Startup, Sep 2023 \- Dec 2023 · 4 mos, San Francisco Bay Area, Engineered a churn decisioning pipeline achieving 85% accuracy by explicitly optimizing the trade\-off between false positives and customer retention cost, enabling targeted, cost\-aware intervention strategies\. Full\-stack Developer at ARTIFICIAL INTELLIGENCE TOOLS AND APPLICATIONS, May 2023 \- Jun 2023 · 2 mos, Prayagraj, Uttar Pradesh, India, • Developed and Launched a Dynamic website providing a seamless platform for students\. • Implemented an immersive user experience, enabling easy navigation of program details, curriculum, and speaker profiles\. • Integrated Google Maps API for precise venue location and streamlined travel arrangements\. • Engineered a secure and efficient online enrollment system, streamlining the registration process for over 55\+ students from diverse colleges across the country\. • Produced 70\+ hours of engaging content to enhance the website’s learning resource\. ## Education Bachelor of Technology \- BTech, Electronics and Communication Engineering, Indian Institute Of Information Technology allahabad [View the original profile](https://tal.club/meet/bhavarth-bhangdia) ```json { "@context": "https://schema.org", "@graph": [ { "@type": "Person", "@id": "https://tal.club/meet/bhavarth-bhangdia#person", "name": "Bhavarth Bhangdia", "url": "https://tal.club/meet/bhavarth-bhangdia", "image": "https://tal.club/meet/bhavarth-bhangdia/profile-photo.jpg", "description": "Meet Bhavarth Bhangdia, Applied AI & Machine Learning Engineer | Generative AI | Agentic Workflows | Scaling Image & Video Models (Diffusion/Temporal) to 6M+ Users | IIIT Allahabad | AI Engineers based in India. AI stack includes antigravity, cursor, and opencode.", "jobTitle": "Machine Learning Engineer", "worksFor": { "@type": "Organization", "name": "Pixlr Online AI Photo Editor" }, "hasOccupation": { "@type": "Occupation", "name": "Machine Learning Engineer" }, "homeLocation": { "@type": "Place", "name": "India" }, "alumniOf": [ { "@type": "EducationalOrganization", "name": "Indian Institute Of Information Technology allahabad" } ], "knowsAbout": [ "antigravity", "cursor", "opencode" ], "mainEntityOfPage": { "@id": "https://tal.club/meet/bhavarth-bhangdia#webpage" } }, { "@type": "ProfilePage", "@id": "https://tal.club/meet/bhavarth-bhangdia#webpage", "url": "https://tal.club/meet/bhavarth-bhangdia", "name": "Bhavarth Bhangdia on Tal", "description": "Meet Bhavarth Bhangdia, Applied AI & Machine Learning Engineer | Generative AI | Agentic Workflows | Scaling Image & Video Models (Diffusion/Temporal) to 6M+ Users | IIIT Allahabad | AI Engineers based in India. AI stack includes antigravity, cursor, and opencode.", "dateModified": "2026-08-02T06:38:44.753Z", "isPartOf": { "@id": "https://tal.club/#website" }, "mainEntity": { "@id": "https://tal.club/meet/bhavarth-bhangdia#person" }, "primaryImageOfPage": { "@type": "ImageObject", "url": "https://tal.club/meet/bhavarth-bhangdia/profile-photo.jpg", "contentUrl": "https://tal.club/meet/bhavarth-bhangdia/profile-photo.jpg", "caption": "Bhavarth Bhangdia profile photo on Tal" }, "inLanguage": "en-IN" } ] } ```