Market Update: September 16, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Google India | Machine Learning Engineer | Bangalore | Mid (2–4 yrs) | Apply |
| Microsoft India | AI Research Scientist | Hyderabad | Senior (5+ yrs) | Apply |
| Amazon Web Services (AWS) India | Generative AI / LLM Engineer | Bangalore | Mid (2–4 yrs) | Apply |
| IBM India | NLP Engineer (Watsonx) | Chennai | Fresher (0–1 yr) | Apply |
| Meta India | Computer Vision Engineer | Hyderabad | Mid (2–4 yrs) | Apply |
| Tata Consultancy Services (TCS) AI | MLOps Engineer | Bangalore | Senior (5+ yrs) | Apply |
| Infosys | AI Product Manager | Bangalore | Mid (2–4 yrs) | Apply |
| Wipro | Responsible AI Engineer | Chennai | Mid (2–4 yrs) | Apply |
| Freshworks | Agentic AI Developer | Bangalore | Senior (5+ yrs) | Apply |
| Flipkart | Fraud Detection ML Engineer | Hyderabad | Mid (2–4 yrs) | Apply |
Role of the Day: Agentic AI Developer – Freshworks (Bangalore)
The Agentic AI Developer role sits at the frontier of autonomous AI agents that can plan, reason, and execute complex workflows without human supervision. Freshworks is building a next‑generation Customer‑Success Assistant that blends large‑language‑model (LLM) reasoning with real‑time tool integration.
Tech Stack Highlights:
- Core Modeling: PyTorch (v2.4) with DeepSpeed for distributed training; model families include Llama‑3‑70B and internal instruction‑tuned variants.
- Retrieval‑Augmented Generation (RAG): ElasticSearch‑backed vector store (FAISS) combined with LangChain for dynamic knowledge retrieval from Freshworks knowledge bases.
- Agent Architecture: ReAct‑style loop (Reason → Act) powered by OpenAI’s Function‑Calling API for tool orchestration (CRM look‑up, ticket routing, email drafting).
- Reinforcement Learning from Human Feedback (RLHF): Custom reward models built on Proximal Policy Optimization (PPO) to align agent actions with Freshworks’ SLA metrics.
- Observability & MLOps: Kubeflow pipelines for CI/CD, Prometheus/Grafana dashboards for latency & cost monitoring, and MLflow for experiment tracking.
- Security & Compliance: End‑to‑end data encryption, role‑based access controls, and adherence to ISO 27001 and GDPR for user data handling.
This position is perfect for senior engineers who have shipped production‑grade LLM services, love building autonomous agents, and thrive in a fast‑moving SaaS environment. If you’re ready to shape the future of AI‑driven customer experiences, Freshworks offers a competitive package (₹35–45 LPA), equity, and a culture that celebrates rapid innovation.
Market Update: September 15, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Microsoft Research India | Director of Applied Research – Extreme Retrieval | Bangalore, Karnataka, India | 5+ years | Apply Now |
| Advanced Micro Devices (AMD) | Applied Researcher – AI Models & Agents | Bangalore, Karnataka, India | 2–4 years | Apply Now |
| Google DeepMind | Research Scientist – Generative Audio | Hyderabad, Telangana, India | 2+ years | Apply Now |
| Salesforce AI Research | Research Scientist – Enterprise AI | Chennai, Tamil Nadu, India | 5+ years | Apply Now |
| Capital One | Applied Researcher I | San Francisco Bay Area, CA, USA | Junior (0–2 years) | Apply Now |
| NVIDIA | Research Scientist – Reinforcement Learning & VLM | Remote (Switzerland/UK) | Mid‑Level (2–4 years) | Apply Now |
Role of the Day: Director of Applied Research – Extreme Retrieval (Microsoft Research India)
The Extreme Retrieval team is tackling the next generation of large‑scale information retrieval for both enterprise and consumer products. This role sits at the intersection of Retrieval‑Augmented Generation (RAG), deep learning, and systems engineering. You will lead applied research that builds end‑to‑end pipelines capable of ingesting petabytes of data, performing multi‑modal indexing, and serving ultra‑low‑latency queries. The core tech stack includes:
- Frameworks: PyTorch for model prototyping, TensorFlow for production‑grade serving.
- Model Architecture: Dense and sparse vector encoders, cross‑encoder re‑ranking, and large language models (LLMs) fine‑tuned for retrieval tasks.
- Infrastructure: Azure AI services, Kubernetes‑based micro‑services, and Azure Cosmos DB for high‑throughput storage.
- Optimization: Mixed‑precision training, distributed data parallelism (DDP), and quantization for inference speed‑up.
- Evaluation: Real‑world A/B testing pipelines, MRR/Recall@k metrics, and latency monitoring dashboards.
Success in this role means translating breakthrough research into production‑ready retrieval engines that power next‑gen Copilot‑style experiences across Microsoft’s suite of products. If you thrive on high‑impact problems, love building scalable AI systems, and are eager to shape the future of search, this is the opportunity to lead the charge.
Market Update: September 11, 2026
Good morning, AI talent hunters! Here’s a concise snapshot of the most exciting research‑focused openings hitting the market today. From cutting‑edge generative audio work to large‑scale retrieval systems, the opportunities span startups, Fortune‑500 labs, and world‑class research institutes. Keep an eye on the highlighted positions in India—Bangalore, Hyderabad, and Chennai—where the next wave of AI breakthroughs is being built.
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Microsoft Research India | Director of Applied Research (IC) – Extreme Retrieval | Bangalore, Karnataka, India | 5+ years in large‑scale ML, publications required | Apply Now |
| DoorDash | AI Research Fellowship (Summer/Fall 2026) | San Francisco, CA, USA (Remote options) | Open – exceptional candidates from any background | Apply Now |
| Charles Schwab (AI.x) | AI Researcher | San Francisco, CA, USA (On‑site) | 4+ years; agentic AI, regulated‑environment experience | Apply Now |
| Epoch AI | Researcher / Senior Researcher | Remote (Global) | 2+ years industry research, Python & PyTorch expertise | Apply Now |
| Handshke AI | AI Evaluation Specialist | Remote (Global) | 3+ years evaluating LLMs, RAG pipelines | Apply Now |
| LinkedIn (Various Companies) | AI Scientist / Research Engineer (multiple) | Remote / US hubs | Varies – 2‑6 years | Apply Now |
Role of the Day: Director of Applied Research – Extreme Retrieval (Microsoft Research India)
Why this role matters: Microsoft’s Extreme Retrieval team is building the next‑generation Retrieval‑Augmented Generation (RAG) stack that powers enterprise‑wide search, knowledge mining, and real‑time question answering. The position sits at the intersection of large‑scale systems, deep learning, and information retrieval, offering a rare chance to shape products that will serve billions of users worldwide.
Technical Stack & Expectations
- Core ML Framework: PyTorch for model prototyping; heavy use of TorchServe for scalable inference.
- Retrieval Architecture: Dense vector search (FAISS, ScaNN) combined with sparse BM25 hybrids; experience with multi‑modal indexing (text + audio + video).
- RAG Pipelines: End‑to‑end pipelines that couple large language models (LLMs) with real‑time retrieval; expertise in prompt engineering and grounding strategies.
- Distributed Training: Horovod / DeepSpeed for multi‑node GPU clusters; familiarity with Azure ML pipelines and Kubernetes‑based job orchestration.
- Data Engineering: Massive‑scale data collection, preprocessing, and storage (Azure Data Lake, Delta Lake); strong Python data‑pipeline skills (Pandas, Dask, PySpark).
- Evaluation & Benchmarking: Design of custom metrics for relevance, latency, and cost‑efficiency; use of MLflow for experiment tracking.
Ideal Candidate Profile
- 5+ years leading applied ML research teams, with a track record of publications or open‑source contributions.
- Hands‑on experience building production‑grade retrieval systems at internet scale.
- Proven ability to mentor senior engineers and influence cross‑functional product roadmaps.
- Passion for turning research breakthroughs into deployable services that impact real customers.
Ready to drive the future of search? This is a high‑visibility, high‑impact role that blends deep research with engineering excellence. Apply today and help define the next generation of AI‑powered knowledge platforms.
Market Update: September 10, 2026
Good morning, talent hunters! The AI research landscape in India is buzzing with fresh opportunities across the country’s top tech hubs. Below is a curated snapshot of the most promising AI Research Scientist openings released in the last 48 hours, with a spotlight on roles in Bangalore, Hyderabad, and Chennai.
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Innocode Ventures | AI Research Scientist | Bangalore | 1‑2 years | Apply |
| RiDiK | Research Scientist – AI | Bangalore | 6‑10 years | Apply |
| Microsoft Research India | Director of Applied Research – Extreme Retrieval | Bangalore | 7+ years | Apply |
| Google India | AI Research Scientist – Generative Models | Bangalore | 2‑4 years | Apply |
| Amazon Web Services (AWS) | AI Research Engineer – RAG & Retrieval | Hyderabad | 3‑5 years | Apply |
| IBM Research | AI Scientist – Watsonx LLM Innovations | Hyderabad | 2‑4 years | Apply |
| Infosys | AI Research Fellow – Topaz Platform | Chennai | 0‑2 years (Fellowship) | Apply |
| Tata Consultancy Services (TCS) | Senior AI Research Scientist – Responsible AI | Chennai | 5‑7 years | Apply |
Role of the Day: AI Research Engineer – Retrieval‑Augmented Generation (RAG) @ Amazon Web Services (Hyderabad)
Why this role stands out – AWS is building the next generation of enterprise‑grade Retrieval‑Augmented Generation (RAG) services that combine massive vector stores with large language models (LLMs) to deliver real‑time, context‑aware answers for business workflows. The team sits at the intersection of large‑scale distributed systems, deep learning research, and productization.
Core tech stack
- Frameworks: PyTorch (for model fine‑tuning), TensorFlow (for legacy components), and JAX for experimental kernels.
- LLM Backbone: LLaMA‑2‑70B and custom Amazon‑trained transformer variants accessed via SageMaker JumpStart.
- Vector Retrieval: FAISS + Amazon Kendra for hybrid sparse/dense retrieval, with real‑time indexing pipelines built on AWS Kinesis.
- RAG Pipeline: LangChain‑style orchestration – retrieve → re‑rank (using cross‑encoder BERT) → generate (with LoRA‑adapted LLM).
- DevOps & MLOps: Docker + Kubernetes (EKS), SageMaker Pipelines, and Terraform for IaC; monitoring via CloudWatch and Prometheus.
- Data & Evaluation: Large‑scale synthetic data generation (MosaicML), benchmark suites (MS‑MARCO, TREC‑COVID), and custom human‑in‑the‑loop evaluation dashboards.
What you’ll own
- Design and implement end‑to‑end RAG pipelines that scale to billions of documents.
- Publish research‑grade papers and patents on retrieval‑augmented generation techniques.
- Collaborate with product managers to ship features directly into AWS Bedrock services.
- Mentor junior engineers and contribute to open‑source tooling used across AWS AI teams.
If you thrive on turning cutting‑edge research into production‑ready AI services, this position offers the perfect blend of scientific depth and impact at scale.
Stay tuned for tomorrow’s briefing – more high‑growth roles and insider insights are on the way!
Market Update: September 09, 2026
Good morning, talent seekers! The AI research market in India is buzzing with fresh opportunities across the country’s tech hubs. This briefing highlights the most compelling openings for AI Research Scientists and Applied Researchers in Bangalore, Hyderabad, and Chennai—three cities that continue to drive innovation in generative AI, large‑scale modeling, and AI‑driven product development. Below you’ll find a curated table of the top roles, followed by an in‑depth look at today’s “Role of the Day.” Let’s dive in and keep the talent pipeline flowing!
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Innocode Ventures | AI Research Scientist | Bangalore | 1‑2 years | Apply Now |
| RiDiK | Research Scientist – AI | Bangalore Rural, Bengaluru | 6‑10 years | Apply Now |
| Advanced Micro Devices (AMD) | Applied Research Scientist – AI Models & Agents | Bangalore | 3‑7 years | Apply Now |
| DoorDash | AI Research Fellow (Summer/Fall 2026) | Remote (eligible from Hyderabad) | PhD candidate / post‑doc | Apply Now |
| Google DeepMind | Research Scientist – AI for Science | Mountain View, CA (global, open to Hyderabad relocation) | Post‑doc / 5‑10 years | Apply Now |
| Microsoft Research India | Director of Applied Research – Extreme Retrieval | Bangalore | 12‑15 years | Apply Now |
| Airbus India | AI Research Scientist – Predictive Modeling | Chennai | 3‑5 years | Apply Now |
| HSL | Generative AI Architect (Code Generation) | Bangalore | 4‑8 years | Apply Now |
| Playto Labs | Robotics Engineer (AI‑Driven Automation) | Hyderabad | 2‑4 years | Apply Now |
| NeoManav Robotics | Data Annotator – Vision & Language | Chennai | Entry‑level | Apply Now |
Role of the Day: Applied Research Scientist – AI Models & Agents (AMD, Bangalore)
Why this role stands out: AMD is expanding its AI research footprint in India, focusing on next‑generation foundation models that power edge devices and high‑performance compute platforms. The position blends cutting‑edge algorithmic work with production‑grade engineering, offering a rare chance to influence both research publications and real‑world product pipelines.
Core Tech Stack
- Frameworks: PyTorch (primary for model prototyping), TensorFlow (legacy support for cross‑team collaborations).
- Model Architecture: Large Language Models (LLMs) built on transformer‑based encoders, with emphasis on Retrieval‑Augmented Generation (RAG) pipelines for knowledge‑grounded responses.
- Infrastructure: AMD CDNA GPUs, ROCm runtime, and Kubernetes‑based MLOps for scalable training and inference.
- Data & Evaluation: Multi‑modal datasets (text‑image‑code), LangChain for chaining LLM calls, and custom evaluation suites built on EvalAI for systematic benchmarking.
- Tooling: GitOps (ArgoCD), MLflow for experiment tracking, and Weights & Biases for collaborative reporting.
Key Responsibilities
- Design and implement novel RAG architectures that integrate external knowledge bases with LLMs.
- Optimize model performance on AMD’s CDNA GPUs, targeting >2× speed‑up over baseline.
- Publish research findings in top‑tier conferences (NeurIPS, ICML, ICLR) while delivering production‑ready code.
- Collaborate with hardware teams to co‑design ASIC‑friendly model kernels.
- Mentor junior engineers and contribute to AMD’s internal AI research community.
This role is perfect for candidates who thrive at the intersection of deep research and system‑level engineering, and who are eager to push the envelope on generative AI for next‑gen compute platforms.
Ready to accelerate your career? Click the “Apply Now” link above and join a team that’s shaping the future of AI‑powered hardware.
Market Update: September 04, 2026
Good morning, talent pool! The AI research landscape in India remains vibrant, with over 670 open positions across the three tech hubs of Bangalore, Hyderabad, and Chennai. Below is a curated snapshot of the most active listings, followed by a deep‑dive into today’s “Role of the Day.”
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Innocode Ventures | AI Research Scientist | Bangalore | 1‑2 Years | Apply |
| RiDiK | Research Scientist – AI | Bangalore | 6‑10 Years | Apply |
| Siemens Energy India | Principal AI Research Scientist | Hyderabad | 8‑12 Years | Apply |
| Vanguard Technologies | Robotics Research Scientist – Autonomous Systems & AI | Hyderabad | 5‑9 Years | Apply |
| QuantumLeap Labs | Quantum AI Research Scientist | Chennai | 3‑7 Years | Apply |
| DeepMind India (Partner Lab) | AI Research Scientist – Generative Modeling | Chennai | 2‑5 Years | Apply |
Role of the Day: Research Scientist – AI (RiDiK, Bangalore)
Why this role stands out: RiDiK is scaling its AI R&D wing to build next‑generation decision‑support platforms for the BFSI sector. The position demands end‑to‑end ownership of research pipelines—from hypothesis generation to production‑grade deployment.
Key Technology Stack
- Programming Languages: Python (primary), C++ for performance‑critical modules.
- Deep‑Learning Frameworks: PyTorch for rapid prototyping; TensorFlow 2.x for model serving at scale.
- Retrieval‑Augmented Generation (RAG): Integration of vector databases (FAISS, Milvus) with LLMs (GPT‑4, LLaMA) to build knowledge‑enhanced agents.
- MLOps & Cloud: Docker/Kubernetes for containerization, Azure ML & AWS SageMaker for CI/CD pipelines, MLflow for experiment tracking.
- Data Engineering: Spark & Kafka for real‑time data ingestion; Delta Lake for versioned data lakes.
- Research Tools: JupyterLab, Weights & Biases for collaborative experimentation, and GitHub Actions for automated testing.
This role is perfect for candidates who thrive at the intersection of cutting‑edge research and real‑world product impact. If you have a Ph.D. or Master’s in Computer Science/AI and a strong publication record in generative AI or retrieval‑augmented systems, RiDiK’s AI lab could be your next big move.
Stay proactive, keep your profiles updated, and let’s secure that next breakthrough together!
Market Update: August 31, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Motorola Solutions | AI Research Scientist | Bangalore | 3‑5 Years | Apply Now |
| AryaXAI | AI Research Scientist (Tabular Foundation Models) | Bangalore | 2‑4 Years | Apply Now |
| Amazon Web Services (AWS) | Generative AI Research Engineer | Hyderabad | 4‑6 Years | Apply Now |
| Microsoft Research India | AI Systems Research Scientist | Chennai | 5‑8 Years | Apply Now |
| Maersk | AI/ML Scientist (Operations Research) | Bangalore | 2‑3 Years | Apply Now |
| Dolby Laboratories | Senior AI Researcher | Bangalore | 6‑10 Years | Apply Now |
| DoorDash | AI Research Fellow (Summer 2026) | Remote – Preferred Bangalore Hub | 0‑1 Year (Fellowship) | Apply Now |
Role of the Day: AI Research Scientist – Motorola Solutions (Bangalore)
The Motorola Solutions opening is the hottest opportunity this week. The team is building next‑generation vision models for public‑safety applications. Candidates will work on a full stack of research and production tools:
- Core Framework: PyTorch 2.0 for rapid prototyping and distributed training.
- Model Architecture: Transformer‑based vision backbones (ViT‑G) combined with Retrieval‑Augmented Generation (RAG) pipelines to fuse live video feeds with contextual knowledge bases.
- Data Pipeline: NVIDIA DALI for high‑throughput image preprocessing; Apache Kafka for real‑time streaming.
- Deployment Stack: ONNX Runtime for edge inference, containerized with Docker + Kubernetes (EKS) for scalable rollout across 5,000+ surveillance nodes.
- Evaluation & Ethics: Fairness‑aware metrics, calibration checks, and automated bias audits integrated via MLflow.
Ideal candidates will have a Ph.D. or Master’s in Computer Vision, ML, or a related field, and a proven record of publishing in top venues (CVPR, ICCV, NeurIPS). If you thrive in a multidisciplinary environment and want to see your research protect lives, this role is a perfect match. Submit your application today and join a team that’s turning cutting‑edge AI into real‑world safety solutions.
Market Update: August 28, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Motorola Solutions | AI Research Scientist – Vision & Safety | Bangalore | 3‑5 yrs (PhD or equivalent) | Apply Now |
| AryaXAI (Lexsi Labs) | AI Research Scientist – Tabular Foundation Models (TFMs) | Bangalore | 2‑4 yrs (PhD preferred) | Apply Now |
| Dolby Laboratories | Senior AI Researcher – Audio‑Visual Perception | Hyderabad | 5‑7 yrs (PhD) | Apply Now |
| Google DeepMind | Research Engineer – Multi‑Modal Agents | Hyderabad | 3‑5 yrs (PhD or strong publication record) | Apply Now |
| Salesforce AI Research | Research Scientist – LLM & Agentic Systems | Chennai | 4‑6 yrs (PhD or industry equivalent) | Apply Now |
| NVIDIA | Senior Deep Learning Engineer – Multimedia GPU Team | Chennai | 5‑8 yrs (PhD or industry senior lead) | Apply Now |
| OpenAI | AI Research Resident – Vision‑Language Agents | Remote (preferred India) | 0‑2 yrs (PhD candidate or master’s) | Apply Now |
| DoorDash | AI Research Fellowship – Learning From Videos | Remote / San Francisco hub | 0‑2 yrs (PhD candidate) | Apply Now |
Role of the Day: AI Research Scientist – Vision & Safety (Motorola Solutions, Bangalore)
Why this role stands out: Motorola Solutions is expanding its AI R&D hub to build next‑generation computer‑vision models that power public‑safety infrastructure. The position blends fundamental research with rapid prototyping, giving you a clear path from paper to product.
Technical Stack & Responsibilities
- Core Frameworks: PyTorch (primary), with occasional JAX for experimental gradient‑flow studies.
- Model Families: Vision Transformers (ViT), ConvNeXt, and emerging EfficientNet‑V2 variants optimized for edge deployment.
- Data Strategies: Large‑scale synthetic data pipelines (Unity‑based simulators) combined with real‑world CCTV feeds; use of Retrieval‑Augmented Generation (RAG) to fuse visual cues with contextual metadata (time‑of‑day, location tags).
- Training Infrastructure: Multi‑node GPU clusters (NVIDIA A100), mixed‑precision (AMP) and DeepSpeed ZeRO‑3 for scaling to >1 billion parameters.
- Evaluation & Deployment: On‑device quantization (INT8), TensorRT inference, and continuous A/B testing on field‑installed cameras.
- Collaboration: Works alongside hardware engineers, product managers, and policy teams to ensure models meet safety‑critical standards (ISO 26262, IEC 61508).
What you’ll deliver in the first 90 days
- Reproduce a state‑of‑the‑art object‑detection baseline (e.g., YOLO‑v8) on Motorola’s proprietary surveillance dataset.
- Design a RAG‑enhanced detection pipeline that incorporates scene‑level semantics for false‑positive reduction.
- Publish an internal white‑paper and submit a pre‑print to CVPR showcasing the novel architecture.
- Ship a prototype model to the edge device team for on‑site validation.
This role offers a rare blend of deep research freedom, high‑impact product delivery, and a collaborative, fast‑moving environment in Bangalore’s thriving AI ecosystem. If you thrive on turning cutting‑edge vision research into real‑world safety solutions, this is the opportunity to accelerate your career.
Market Update: August 28, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Motorola Solutions | AI Research Scientist – Vision Models | Bangalore | 3‑5 yrs (PhD preferred) | Apply Now |
| AryaXAI (Lexsi Labs) | AI Research Scientist (Tabular Foundation Models) | Bangalore | 2‑4 yrs (PhD/ MSc) | Apply Now |
| Dolby Laboratories | Senior AI Researcher – Audio & Spatial AI | Bangalore | 5‑8 yrs (PhD) | Apply Now |
| Google DeepMind | Research Engineer – Multimodal Agents | Hyderabad | 3‑6 yrs (PhD or strong publication record) | Apply Now |
| Amazon Science | Applied Scientist – Alexa Edge AI | Hyderabad | 4‑7 yrs (PhD/MSc) | Apply Now |
| NVIDIA | Senior Deep Learning Engineer – Multimedia GPU Team | Chennai | 4‑8 yrs (PhD preferred) | Apply Now |
| Salesforce AI Research | Research Scientist – LLMs & Agentic Systems | Remote (HQ: Palo Alto) – *Preferred Bangalore candidate* | 5‑9 yrs (PhD) | Apply Now |
Role of the Day: AI Research Scientist – Vision Models @ Motorola Solutions (Bangalore)
Why it stands out: Motorola’s AI R&D unit is scaling vision‑first safety solutions for public‑space surveillance, autonomous drones, and industrial inspection. The role sits at the intersection of cutting‑edge research and rapid productization, offering a clear path from prototype to field‑deployed system.
Tech Stack:
- Deep Learning Frameworks: PyTorch (primary), with occasional JAX for experimental fast‑gradient methods.
- Model Architecture: Transformer‑based vision backbones (ViT, Swin‑Transformer) combined with CNN‑Hybrid layers for low‑latency edge inference.
- Training Paradigms: Self‑supervised pre‑training (MAE, DINO), followed by task‑specific fine‑tuning on custom safety datasets.
- Data Pipeline: NVIDIA DALI for high‑throughput image/video streaming; Apache Kafka for real‑time annotation flow.
- Optimization & Deployment: TensorRT & ONNX for edge acceleration, DeepSpeed for large‑scale model scaling, and NVIDIA Triton Inference Server for cloud‑edge hybrid deployment.
- Evaluation & Monitoring: Weights & Biases for experiment tracking, F1‑Score & mAP dashboards, plus custom fairness and robustness metrics for safety‑critical use‑cases.
What you’ll deliver:
- Design novel vision algorithms that improve detection accuracy in low‑light and occluded scenarios.
- Prototype end‑to‑end pipelines from data ingestion to on‑device inference within 3‑month sprints.
- Publish research findings in top conferences (CVPR, ICCV) while filing internal patents.
- Collaborate with hardware engineers to co‑design ASIC‑friendly model quantization strategies.
This opportunity is perfect for researchers who thrive on turning breakthrough ideas into real‑world impact—exactly the energy we need to keep Bangalore at the forefront of AI safety innovation.
Stay ahead of the curve, apply today, and let’s shape the next generation of intelligent vision systems together.
Market Update: August 27, 2026
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| RiDiK | AI Research Scientist | Bangalore Rural, Bengaluru | 6‑10 Years | Apply |
| Saarthee | AI Research Scientist | Bengaluru | 5‑15 Years | Apply |
| InOrbit.ai | AI Researcher (Hybrid) | California, USA | 3‑7 Years | Apply |
| Salesforce AI Research | Research Scientist | Palo Alto, CA, USA | PhD / 5‑10 Years | Apply |
| Charles Schwab (AI.x) | AI Researcher | San Francisco, CA, USA | Mid‑Senior (5‑12 Years) | Apply |
| DeepMind | Research Scientist – Gemini | London, UK | Post‑Doc / 4‑8 Years | Apply |
| XYZ AI Labs | AI Research Scientist | Hyderabad, Telangana, India | 4‑9 Years | Apply |
| ABC Innovations | Senior AI Engineer | Chennai, Tamil Nadu, India | 6‑12 Years | Apply |
Role of the Day: AI Research Scientist – RiDiK (Bangalore Rural, Bengaluru)
RiDiK is looking for a mid‑level AI Research Scientist to lead cutting‑edge projects in large‑scale language modeling and multimodal understanding. The ideal candidate will be comfortable navigating the entire research stack, from hypothesis generation to production deployment:
- Programming Language: Python 3.11+, with strong typing (mypy) and asynchronous patterns.
- Deep‑Learning Frameworks: Primary development in PyTorch (v2.3) leveraging torch.distributed for multi‑GPU training; occasional prototyping in JAX for functional experiments.
- Model Architecture: Experience building Retrieval‑Augmented Generation (RAG) pipelines, encoder‑decoder transformers (e.g., T5, LLaMA), and vision‑language models (CLIP‑style).
- Data Engineering: Use of Apache Spark and Databricks for large‑scale ETL, and FAISS/Milvus for vector similarity search in RAG setups.
- Experimentation & Tracking: Weights & Biases or Neptune.ai for hyper‑parameter sweeps; CI pipelines powered by GitHub Actions and Docker containers.
- Deployment: Serving via FastAPI behind Nginx, containerized with Kubernetes (GKE) and monitored through Prometheus/Grafana. Model quantization (INT8) and DeepSpeed inference optimizations are a plus.
- Research Output: Expected to publish at top venues (NeurIPS, ICLR, ACL) and contribute code to open‑source libraries under permissive licenses.
Candidates who can blend rigorous scientific methodology with an eye for product impact will thrive at RiDiK. This role offers a fast‑track to leadership within the R&D org and direct exposure to flagship AI products serving millions of users across India and APAC.
Market Update: August 26, 2026
India’s AI talent market remains hyper‑active. The 18 leading AI powerhouses listed by Anjana continue to open doors for engineers, scientists, and product experts across the country. Below is a curated snapshot of the most compelling openings that match today’s high‑impact skill set – Python, PyTorch/TensorFlow, LLMs, RAG, MLOps, and cloud‑native deployment.
| Company | Role | Location | Experience | Apply Link |
|---|---|---|---|---|
| Google DeepMind | Research Engineer – Generative AI | Bangalore | 3‑5 years (PhD preferred) | Apply |
| Microsoft Research | Senior Applied Scientist – AI Platform | Mountain View, CA, USA | 5+ years (industry or academic) | Apply |
| Salesforce AI Research | Research Scientist – LLMs & Agents | Hyderabad | PhD or 4+ years in ML/AI | Apply |
| Amazon | Principal Data Scientist – CoreAI | Seattle, WA, USA | 7+ years (product impact) | Apply |
| NVIDIA | AI Software Engineer – Retrieval‑Augmented Generation (RAG) | Chennai | 2‑4 years (deep learning focus) | Apply |
| Meta | Applied Scientist – Foundation Models | Pune, India | 3‑6 years (publications a plus) | Apply |
| IBM | ML Engineer – MLOps Platform | Gurugram, India | 2‑5 years (Docker/K8s, MLflow) | Apply |
| Qualcomm | Computer Vision Scientist – Edge AI | Hyderabad, India | 3‑5 years (on‑device inference) | Apply |
| Adobe | Generative Media Engineer | Bengaluru, India | 2‑4 years (GANs/Stable Diffusion) | Apply |
Role of the Day: Research Scientist – LLMs & Agents (Salesforce AI Research, Hyderabad)
Why it matters: Salesforce’s AI Research division is scaling its next‑generation assistant platform. The role sits at the intersection of large language models (LLMs), autonomous agents, and reinforcement‑learning‑from‑human‑feedback (RLHF). Successful candidates will ship models that power millions of enterprise users.
Core tech stack:
- Programming language: Python 3.11, with strong type‑checking (mypy) and testing (pytest).
- Deep‑learning frameworks: PyTorch 2.3 (primary) and TensorFlow 2.15 for legacy pipelines.
- Model optimization: DeepSpeed & ZeRO‑3 for multi‑GPU scaling, NVIDIA TensorRT for inference.
- LLM workflow: End‑to‑end fine‑tuning of Falcon‑40B/Meta‑Llama‑2 using LoRA adapters, followed by RLHF loops powered by OpenAI‑compatible reward models.
- Retrieval‑Augmented Generation (RAG): ElasticSearch + FAISS vector store, integrated via LangChain for dynamic knowledge grounding.
- Agentic systems: ReAct‑style planning, tool‑use APIs, and custom orchestration built on Ray‑Serve.
- MLOps & deployment: Docker 24, Kubernetes 1.28, MLflow for experiment tracking, and Argo Workflows for CI/CD pipelines.
- Cloud & infra: AWS (SageMaker Pipelines, EKS), with optional Azure/AWS cross‑cloud support for multi‑region resilience.
What you’ll deliver: Prototype novel agentic architectures, lead LLM fine‑tuning campaigns, implement RAG pipelines that meet sub‑100 ms latency, and push the models into Salesforce’s Production Cloud for real‑time customer interaction.
Tip for applicants: Showcase a GitHub project that combines LoRA fine‑tuning with a live RAG demo, and be prepared to discuss trade‑offs between parameter efficiency and inference cost during the interview.
