New Application Development & Demo Deployment Lead

Saaki-Argus-and-Averil-Consulting; Bangalore Rural; On-site; INR 3,500,000 - 7,000,000; Full time

Job summary

Saaki, Argus and Averil Consulting seeks a hands-on AI/HPC application engineer to port workloads such as LLMs and molecular dynamics to our accelerator platform and to build live demos for customers.

You will lead porting efforts, develop reproducible benchmarks, present results, and support SDK onboarding for customer proofs of concept. Travel to customer demos and conferences will be required.

Qualifications

• B.Tech/B.Tech or M.S. in CS, Electrical Engineering or related field.
• 8–12 years in HPC/AI application engineering, accelerator software or platform bring-up.
• Hands-on porting experience using CUDA, OpenCL, HIP or MPI.
• Strong Python and C/C++; PyTorch and LLM inference experience.

Responsibilities

• Lead porting of AI and HPC applications, including LLMs, molecular dynamics, CFD and genomics.
• Build, benchmark and present live demos.
• Develop a fair benchmarking methodology covering bandwidth, throughput, power and accuracy.
• Support customer evaluations and SDK onboarding; document results.

Skills

CUDA/OpenCL/HIP/MPI

Python & C/C++

PyTorch & LLM inference

Performance profiling

Technical communication

Travel willingness

Education

B.Tech/M.Tech or M.S. in CS/EE or related

Tools

GROMACS

OpenFOAM

BWA

BLAST

Job description

About the roleWere building next-generation AI and HPC acceleration technology and need a hands-on engineer to make real applications run on it. Youll port AI and scientific workloads to our new silicon, build live demos, and support customer proofs of concept.What youll doLead porting of AI and HPC applications, including LLMs, molecular dynamics, CFD and genomicsBuild, benchmark and present live demosDevelop a fair, reproducible benchmarking methodology covering bandwidth, throughput, power and accuracySupport customer evaluations and SDK onboarding, and document resultsEscalate SDK gaps and performance bottlenecks to the platform team with reproducible test casesWhat were looking forB.Tech/M.Tech or M.S. in CS, Electrical Engineering or a related field8-12 years in HPC/AI application engineering, accelerator software or platform bring-upHands-on porting experience using CUDA, OpenCL, HIP or MPIStrong Python and C/C++; PyTorch and LLM inference experienceExperience with roofline analysis, memory bandwidth measurement and latency profilingClear communicator with engineering and non-technical audiencesWillingness to travel for customer demos, PoCs and conferencesNice to haveLLM inference frameworks (vLLM, Hugging Face, llama.cpp) on non-NVIDIA hardwareExperience with GROMACS, OpenFOAM, BWA or BLASTBackground at an AI accelerator startup or HPC vendor