Research Engineer - Multimodal AI, Data Quality & Model Evaluation
- Discuss this with your agent
- Open in Claude
- Open in ChatGPT
Prepare application
- Draft this with your agent
- Open in Claude
- Open in ChatGPT
Role details
Tech stack
+10 more
Job description
PySpark Research Operations Scalability Reliability Transcribing Communication Data Analysis Data Ingestion Computer Vision Language Models Safety Assurance Machine Learning Business Valuation Regression Testing Course Evaluations Workflow Management Full Stack Development Business Transformation Audio Signal Processing SQL (Programming Language) Fine-tuning (Deep Learning) Python (Programming Language) Artificial Intelligence Workflow, This fully remote Research Engineer role focuses on improving the data, evaluation systems, and production workflows behind applied AI models. You will work across research and engineering, improving dataset quality, investigating evaluation failures, and helping move experimental AI workflows into production., * Analyze and improve the quality, coverage, consistency, and reliability of AI training and evaluation datasets.
- Perform failure analysis to determine why datasets, pipelines, models, or evaluations are underperforming.
- Build evaluation, QA, validation, regression-testing, monitoring, and reporting pipelines.
- Support datasets involving speech, audio, video, computer vision, image-to-text, translation, or multimodal AI.
- Develop scalable Python and SQL workflows for data ingestion, transformation, filtering, enrichment, validation, and delivery.
- Support model training, fine-tuning, post-training, evaluation, and synthetic-data workflows.
- Partner with research scientists and engineers to translate experimental requirements into scalable technical solutions.
- Troubleshoot data inconsistencies, pipeline failures, model regressions, and performance bottlenecks.
- Help transition research-focused datasets and workflows into reliable production systems.
- Operate independently and proactively bring solutions and recommendations in ambiguous environments.
Requirements
The ideal candidate has production experience with speech, audio, video, computer vision, image-to-text, or multimodal AI, along with enough engineering depth to build reliable data and evaluation pipelines. This is a hands-on role for someone comfortable working with large, imperfect datasets and solving ambiguous problems independently., * Production experience in at least one relevant domain: speech, audio, video, computer vision, image-to-text, translation, OR multimodal AI.
- Experience supporting model training, fine-tuning, post-training, synthetic-data, or model-evaluation workflows.
- Hands-on experience analyzing and improving the quality of training or evaluation datasets.
- Ability to identify whether failures originate from datasets, labels, features, pipelines, model behavior, or evaluation methodology.
- Experience building evaluation, QA, validation, regression-testing, or monitoring pipelines.
- Strong Python experience supporting applied ML or research workflows.
- Experience working with large, incomplete, inconsistent, or imperfect real-world datasets.
- Industry experience partnering with research scientists, ML engineers, data engineers, or cross-functional stakeholders.
- Ability to bridge research and engineering rather than operating solely within an academic environment.
- Strong communication skills and ability to work autonomously when requirements are not fully defined.
Nice to Haves:
- Experience with SFT, LoRA/QLoRA, DPO, RLHF, GRPO, reward modeling, preference data, human feedback, or knowledge distillation.
- Experience generating, curating, filtering, or validating synthetic training data.
- Experience with ASR, TTS, multilingual speech data, audio processing, transcription evaluation, or voice AI.
- Experience with vision-language models, video understanding, image-text datasets, or multimodal foundation models.
- Experience with distributed processing frameworks such as Spark, PySpark, Ray, or Hadoop.
- Masters degree or equivalent relevant industry experience preferred.
- Publications, patents, or conference experience are a plus but not required., Individual compensation offered for this position within this range will depend on many factors, including qualifications, skills, relevant experience, job knowledge, geographic location, internal equity, and other pertinent job-related factors.
About the company
Weβre partners in transformation. We help clients activate ideas and solutions to take advantage of a new world of opportunity. We are a team of 80,000 strong, working with over 6,000 clients, including 80% of the Fortune 500, across North America, Europe and Asia. As an industry leader in Full-Stack Technology Services, Talent Services, and real-world application, we work with progressive leaders to drive change. Thatβs the power of true partnership. TEKsystems is an Allegis Group company., Were a leading provider of business and technology services. We accelerate business transformation for our customers. Our expertise in strategy, design, execution and operations unlocks business value through a range of solutions. Were a team of 80,000 strong, working with over 6,000 customers, including 80% of the Fortune 500 across North America, Europe and Asia, who partner with us for our scale, full-stack capabilities and speed. Were strategic thinkers, hands-on collaborators, helping customers capitalize on change and master the momentum of technology. Were building tomorrow by delivering business outcomes and making positive impacts in our global communities. TEKsystems and TEKsystems Global Services are Allegis Group companies. Learn more at TEKsystems.com.
Apply for this position
This job is hosted externally. Click below to view the full posting and apply.
Prepare application
- Draft this with your agent
- Open in Claude
- Open in ChatGPT
Good distractions
Loading talks and stories from around this roleβ¦