Calibrated Human-in-the-Loop AI Agent Harness (Thesis)

Ericsson

Karlskrona kommun

On-site

SEK 600,000 - 900,000

Full time

5 days ago
Be an early applicant

A complete application in a minute

Tailored resume and cover letter, ready to send.

Get past ATS filters

Job summary

Ericsson is seeking a researcher to design an extensible agent harness that improves collaboration between humans and AI agents. You will create inline clarifications, confidence indicators, and a decision-support advisor to balance automation with human review.

The role explores how a shared vocabulary for assumptions and decisions enhances alignment, reduces rework, and keeps humans in control of important judgments. You will prototype in an extensible framework and assess its practical impact.

Qualifications

  • Strong foundations in computer science, information systems, or related fields.
  • Experience with prototyping and evaluating software/AI systems.
  • Interest in AI agents, human-in-the-loop design, or interaction research.
  • Ability to articulate research questions and summarize findings clearly.

Responsibilities

  • Design clarification mechanisms between minimal and explicit questions.
  • Develop an advisor component to assist decision-making and review.
  • Summarize changes, highlight decisions, and identify parts needing review.
  • Explore when to checkpoint vs. autonomous execution based on verification costs.
  • Build a prototype within an extensible agent harness using existing frameworks.
  • Evaluate how the research vocabulary impacts alignment and rework.

Skills

Knowledge of CS
HCI
Software Engineering
Artificial Intelligence
Interdisciplinary research
Communication

Job description

Ericsson is seeking a researcher to design an extensible agent harness that improves collaboration between humans and AI agents. You will create inline clarifications, confidence indicators, and a decision-support advisor to balance automation with human review.

The role explores how a shared vocabulary for assumptions and decisions enhances alignment, reduces rework, and keeps humans in control of important judgments. You will prototype in an extensible framework and assess its practical impact.

Get your free, confidential resume review.

or drag and drop your file here.

Similar jobs

Similar jobs worth comparing

Thesis - Human-in-the-Loop Calibrated Agent Harness
Thesis - Human-in-the-Loop Calibrated Agent Harness

Ericsson • Karlskrona kommun

On-site
SEK 600,000 - 900,000
Chief AI Systems Architect – Autonomous Agent Ecosystem
Chief AI Systems Architect – Autonomous Agent Ecosystem

Ericsson AB • Stockholms kommun

On-site
SEK 1,900,000 - 2,700,000
Lead AI System Architect
Lead AI System Architect

Ericsson AB • Stockholms kommun

On-site
SEK 1,900,000 - 2,700,000
GenAI for Network Config Templates: Authoring & Troubleshooting
GenAI for Network Config Templates: Authoring & Troubleshooting

Ericsson • Karlskrona kommun

On-site
SEK 22,000 - 36,000
Enterprise AI Workflow Architect (Remote Contractor)
Enterprise AI Workflow Architect (Remote Contractor)

YO AI Labs • Göteborgs kommun

Remote
SEK 829,796,000 - 1,659,592,000
AI Engineer: Autonomous Agents, Production-ready
AI Engineer: Autonomous Agents, Production-ready

Epiminds AB • Stockholms kommun

On-site
SEK 900,000 - 1,300,000
Competitive salary and equity
Healthcare and retirement savings
Office catering
Frontier AI Engineer: Multi-Agent Systems & LLMs
Frontier AI Engineer: Multi-Agent Systems & LLMs

Cognizant • Stockholms kommun

Hybrid
SEK 900,000 - 1,500,000
Hybrid work
Career development
Competitive compensation
+1
Agent Engineer, Sweden
Agent Engineer, Sweden

Wonderful Ltd. • Sweden

On-site
SEK 438,740 - 658,111
Thesis: Tracing Input Causality in LLM Agents
Thesis: Tracing Input Causality in LLM Agents

AI Sweden • Göteborgs kommun

Hybrid
SEK 234,000 - 301,000
Hybrid work
Remote work from Lund/Stockholm
Senior AI Engineer
Senior AI Engineer

ALLPS • Stockholms kommun

Hybrid
SEK 900,000 - 1,100,000
Fixed-rate assignments
Freelance or employment through ALLPS
AI interview process
+1