⭐【Python】機械学習モデルの実装、評価、性能改善業務案件⭐

[Python] Machine Learning Model Implementation, Evaluation, and Performance Improvement Project

Source description · Japanese

【JOBID:JA-090298】こんにちは！クラウドワークステック事務局です。「クラウドワークステック」は株式会社クラウドワークス公式運営のフリーランス参画支援サービスです。今回は、【Python】機械学習モデルの実装、評価、性能改善業務案件のお仕事の紹介です。 https://tech.crowdworks.jp/job_offers/97730?ref=pf_request_scout_97730_direct&tracking=pf_request_scout_97730_direct&utm_source=crowdworks&utm_medium=pf_request_scout ✅毎月の稼働時間：96時間程度　（以降、双方合意のもと継続更新予定） ✅報酬：360,000円～468,000円を想定しています　※ご相談可能 ✅勤務形態：フルリモートのお仕事のご案内となります。 ✅備考：業務経験が2年以上ある方を想定しております。 ▼本募集の詳細・応募はこちら（簡単1分） https://tech.crowdworks.jp/job_offers/97730?ref=pf_request_scout_97730_direct&tracking=pf_request_scout_97730_direct&utm_source=crowdworks&utm_medium=pf_request_scout ※人気案件の場合はすぐに募集が終了する恐れがあります。 ※すでに募集が終了している際には、案件の詳細ページではなく、登録用ページが表示されます。 ※システムの都合上、クラウドワークス内のメッセージ返信や応募機能を使ったご相談にはお答えできかねます。大変恐縮ですが、案件へのご質問やご応募は、すべて【上記URL】からお願いいたします。 ---------------- ▼案件応募までの3STEP（簡単1分） STEP1. 下記URLから応募します https://tech.crowdworks.jp/job_offers/97730?ref=pf_request_scout_97730_direct&tracking=pf_request_scout_97730_direct&utm_source=crowdworks&utm_medium=pf_request_scout STEP2.【職務経歴書】を登録します（マイページ＞プロフィール設定＞職務経歴書・ポートフォリオ） → これだけで選考通過率が大幅UP！ STEP3.担当者からの連絡を待ちます ---------------- 上記に限らず、常時600～700件以上の案件がございますので、本案件に限らず、ご興味がある方はぜひご応募をお待ちしております！・ご契約は株式会社クラウドワークスからの業務委託（準委任）契約となります。・本案件は受託企業さまへのお仕事の紹介はできません。あくまで個人のフリーランスとして活動されている方を対象にした案件となります。

Deferred analysis

Implement, evaluate, and improve machine learning models in Python. Full remote, roughly 96 hours per month with continuation subject to mutual agreement. Around 2+ years of business experience expected. Fixed-price listing.

Japan residency: Unknown
Japanese language: Unknown
Japanese company: Unknown
Human days: 10
AI hours: 20
AI input / output: 50000 / 16000 tokens
Version / date: offline-v1 / 2026-10-31T07:09:25.863185+00:00

Evidence, confidence & provider

{"provider": "atmorouter", "model": "atmo/deepseek-v4.1-flash", "confidence": 0.55, "evidence": {"title_en": "【Python】機械学習モデルの実装、評価、性能改善業務案件", "summary_en": "毎月の稼働時間：96時間程度", "human_effort_min_days": "既存の確定済み工数レンジを維持", "human_effort_max_days": "既存の確定済み工数レンジを維持", "ai_effort_hours": "ML model implementation and tuning; code scaffold partly standard but evaluation bespoke.", "ai_input_tokens": "AI支援実行20時間 × 入力2,500トークン/時 の想定（実測ではない）", "ai_output_tokens": "AI支援実行20時間 × 出力800トークン/時 の想定（実測ではない）"}}

Analysis assumptions

{"usage_kind": "estimated", "settings_version": 3, "settings_snapshot": {"version": 3, "model_label": "Sol 6.1 High", "input_usd_per_million": 2.0, "output_usd_per_million": 20.0, "usd_jpy": 157.881601, "human_day_hours": 8.0, "model_hourly_usd": null, "human_hourly_jpy": null, "human_day_jpy": null, "effective_at": "2026-10-30T18:05:00+07:00", "source": "Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).", "approved": 1}, "formula_version": "ai-only-v1", "financial_bounds": {"revenue_min_jpy": 400000.0, "revenue_max_jpy": 500000.0, "profit_min_usd": 2533.12, "profit_max_usd": 3166.51, "profit_min_jpy": 399933.69, "profit_max_jpy": 499933.69}}

Previous validated analyses (5, latest 20)

2026-10-31T07:09:25.863185+00:00

{"source_job_id": "13563683", "title_en": null, "summary_en": null, "japan_resident_required": null, "japanese_language_required": null, "japanese_company_required": null, "human_effort_min_days": 10.0, "human_effort_max_days": 10.0, "ai_effort_hours": 20.0, "ai_input_tokens": 50000, "ai_output_tokens": 16000, "ai_cost_usd": 0.42, "estimated_revenue_jpy": 400000.0, "profit_jpy": 399933.69, "profit_usd": 2533.12, "assumptions_json": "{\"usage_kind\": \"estimated\", \"settings_version\": 3, \"settings_snapshot\": {\"version\": 3, \"model_label\": \"Sol 6.1 High\", \"input_usd_per_million\": 2.0, \"output_usd_per_million\": 20.0, \"usd_jpy\": 157.881601, \"human_day_hours\": 8.0, \"model_hourly_usd\": null, \"human_hourly_jpy\": null, \"human_day_jpy\": null, \"effective_at\": \"2026-10-30T18:05:00+07:00\", \"source\": \"Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).\", \"approved\": 1}, \"formula_version\": \"ai-only-v1\", \"financial_bounds\": {\"revenue_min_jpy\": 400000.0, \"revenue_max_jpy\": 500000.0, \"profit_min_usd\": 2533.12, \"profit_max_usd\": 3166.51, \"profit_min_jpy\": 399933.69, \"profit_max_jpy\": 499933.69}}", "provenance_json": "{\"provider\": \"atmorouter\", \"model\": \"atmo/deepseek-v4.1-flash\", \"confidence\": 0.55, \"evidence\": {\"human_effort_min_days\": \"既存の確定済み工数レンジを維持\", \"human_effort_max_days\": \"既存の確定済み工数レンジを維持\", \"ai_effort_hours\": \"ML model implementation and tuning; code scaffold partly standard but evaluation bespoke.\", \"ai_input_tokens\": \"AI支援実行20時間 × 入力2,500トークン/時 の想定（実測ではない）\", \"ai_output_tokens\": \"AI支援実行20時間 × 出力800トークン/時 の想定（実測ではない）\"}}", "source_hash": "6a515e161e7279064e38e59aa7d0ac4d2dc608643e2298953d857890ef99e93b", "analyzed_at": "2026-10-31T07:08:12.129756+00:00", "analysis_version": "offline-v1"}

2026-10-31T07:08:12.129756+00:00

{"source_job_id": "13563683", "title_en": "[Python] Machine Learning Model Implementation, Evaluation, and Performance Improvement Project", "summary_en": "Implement, evaluate, and improve machine learning models in Python. Full remote, roughly 96 hours per month with continuation subject to mutual agreement. Around 2+ years of business experience expected. Fixed-price listing.", "japan_resident_required": null, "japanese_language_required": null, "japanese_company_required": null, "human_effort_min_days": 10.0, "human_effort_max_days": 10.0, "ai_effort_hours": null, "ai_input_tokens": null, "ai_output_tokens": null, "ai_cost_usd": null, "estimated_revenue_jpy": 400000.0, "profit_jpy": null, "profit_usd": null, "assumptions_json": "{\"settings_version\": 3, \"settings_snapshot\": {\"version\": 3, \"model_label\": \"Sol 6.1 High\", \"input_usd_per_million\": 2.0, \"output_usd_per_million\": 20.0, \"usd_jpy\": 157.881601, \"human_day_hours\": 8.0, \"model_hourly_usd\": null, \"human_hourly_jpy\": null, \"human_day_jpy\": null, \"effective_at\": \"2026-10-30T18:05:00+07:00\", \"source\": \"Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).\", \"approved\": 1}, \"formula_version\": \"ai-only-v1\", \"financial_bounds\": {\"revenue_min_jpy\": 400000.0, \"revenue_max_jpy\": 500000.0, \"profit_min_usd\": null, \"profit_max_usd\": null, \"profit_min_jpy\": null, \"profit_max_jpy\": null}}", "provenance_json": "{\"provider\": \"atmorouter\", \"model\": \"atmo/deepseek-v4.1-flash\", \"confidence\": 0.7, \"evidence\": {\"title_en\": \"【Python】機械学習モデルの実装、評価、性能改善業務案件\", \"summary_en\": \"毎月の稼働時間：96時間程度\", \"human_effort_min_days\": \"毎月の稼働時間：96時間程度（約12人日/月）から推定\", \"human_effort_max_days\": \"毎月の稼働時間：96時間程度（約12人日/月）から推定\"}}", "source_hash": "6a515e161e7279064e38e59aa7d0ac4d2dc608643e2298953d857890ef99e93b", "analyzed_at": "2026-10-31T07:01:42.913745+00:00", "analysis_version": "offline-v1"}

2026-10-31T07:01:42.913745+00:00

{"source_job_id": "13563683", "title_en": "[Python] Machine Learning Model Implementation, Evaluation, and Performance Improvement Project", "summary_en": "Implement, evaluate, and improve machine learning models in Python. Full remote, roughly 96 hours per month with continuation subject to mutual agreement. Around 2+ years of business experience expected. Fixed-price listing.", "japan_resident_required": null, "japanese_language_required": null, "japanese_company_required": null, "human_effort_min_days": 10.0, "human_effort_max_days": null, "ai_effort_hours": null, "ai_input_tokens": null, "ai_output_tokens": null, "ai_cost_usd": null, "estimated_revenue_jpy": 400000.0, "profit_jpy": null, "profit_usd": null, "assumptions_json": "{\"settings_version\": 3, \"settings_snapshot\": {\"version\": 3, \"model_label\": \"Sol 6.1 High\", \"input_usd_per_million\": 2.0, \"output_usd_per_million\": 20.0, \"usd_jpy\": 157.881601, \"human_day_hours\": 8.0, \"model_hourly_usd\": null, \"human_hourly_jpy\": null, \"human_day_jpy\": null, \"effective_at\": \"2026-10-30T18:05:00+07:00\", \"source\": \"Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).\", \"approved\": 1}, \"formula_version\": \"ai-only-v1\", \"financial_bounds\": {\"revenue_min_jpy\": 400000.0, \"revenue_max_jpy\": 500000.0, \"profit_min_usd\": null, \"profit_max_usd\": null, \"profit_min_jpy\": null, \"profit_max_jpy\": null}}", "provenance_json": "{\"provider\": \"atmorouter\", \"model\": \"atmo/deepseek-v4.1-flash\", \"confidence\": 0.7, \"evidence\": {\"title_en\": \"【Python】機械学習モデルの実装、評価、性能改善業務案件\", \"summary_en\": \"毎月の稼働時間：96時間程度\", \"human_effort_min_days\": \"毎月の稼働時間：96時間程度（約12人日/月）から推定\"}}", "source_hash": "6a515e161e7279064e38e59aa7d0ac4d2dc608643e2298953d857890ef99e93b", "analyzed_at": "2026-10-31T07:00:10.525017+00:00", "analysis_version": "offline-v1"}

2026-10-31T07:00:10.525017+00:00

{"source_job_id": "13563683", "title_en": "[Python] Machine Learning Model Implementation, Evaluation, and Performance Improvement Project", "summary_en": "Implement, evaluate, and improve machine learning models in Python. Full remote, roughly 96 hours per month with continuation subject to mutual agreement. Around 2+ years of business experience expected. Fixed-price listing.", "japan_resident_required": null, "japanese_language_required": null, "japanese_company_required": null, "human_effort_min_days": 10.0, "human_effort_max_days": null, "ai_effort_hours": null, "ai_input_tokens": 100000, "ai_output_tokens": 32000, "ai_cost_usd": 0.84, "estimated_revenue_jpy": 400000.0, "profit_jpy": 399867.38, "profit_usd": 2532.7, "assumptions_json": "{\"usage_kind\": \"estimated\", \"settings_version\": 3, \"settings_snapshot\": {\"version\": 3, \"model_label\": \"Sol 6.1 High\", \"input_usd_per_million\": 2.0, \"output_usd_per_million\": 20.0, \"usd_jpy\": 157.881601, \"human_day_hours\": 8.0, \"model_hourly_usd\": null, \"human_hourly_jpy\": null, \"human_day_jpy\": null, \"effective_at\": \"2026-10-30T18:05:00+07:00\", \"source\": \"Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).\", \"approved\": 1}, \"formula_version\": \"ai-only-v1\", \"financial_bounds\": {\"revenue_min_jpy\": 400000.0, \"revenue_max_jpy\": 500000.0, \"profit_min_usd\": 2532.7, \"profit_max_usd\": 3166.09, \"profit_min_jpy\": 399867.38, \"profit_max_jpy\": 499867.38}}", "provenance_json": "{\"provider\": \"atmorouter\", \"model\": \"atmo/deepseek-v4.1-flash\", \"confidence\": 0.7, \"evidence\": {\"title_en\": \"【Python】機械学習モデルの実装、評価、性能改善業務案件\", \"summary_en\": \"毎月の稼働時間：96時間程度\", \"human_effort_min_days\": \"毎月の稼働時間：96時間程度（約12人日/月）から推定\", \"ai_input_tokens\": \"AI支援実行40時間 × 入力2,500トークン/時 の想定（実測ではない）\", \"ai_output_tokens\": \"AI支援実行40時間 × 出力800トークン/時 の想定（実測ではない）\"}}", "source_hash": "6a515e161e7279064e38e59aa7d0ac4d2dc608643e2298953d857890ef99e93b", "analyzed_at": "2026-10-31T05:53:36.231927+00:00", "analysis_version": "offline-v1"}

2026-10-31T05:53:36.231927+00:00

{"source_job_id": "13563683", "title_en": null, "summary_en": null, "japan_resident_required": null, "japanese_language_required": null, "japanese_company_required": null, "human_effort_min_days": null, "human_effort_max_days": null, "ai_effort_hours": null, "ai_input_tokens": null, "ai_output_tokens": null, "ai_cost_usd": null, "estimated_revenue_jpy": 400000.0, "profit_jpy": null, "profit_usd": null, "assumptions_json": "{\"settings_version\": 3, \"settings_snapshot\": {\"version\": 3, \"model_label\": \"Sol 6.1 High\", \"input_usd_per_million\": 2.0, \"output_usd_per_million\": 20.0, \"usd_jpy\": 157.881601, \"human_day_hours\": 8.0, \"model_hourly_usd\": null, \"human_hourly_jpy\": null, \"human_day_jpy\": null, \"effective_at\": \"2026-10-30T18:05:00+07:00\", \"source\": \"Operator-specified (Commander): 2.00 USD input / 20.00 USD output per 1M tokens for Sol 6.1 High, described as the cheapest AtmoRouter pricing. FX USD/JPY 157.881601 from open.er-api.com (2026-10-30T00:02:31Z).\", \"approved\": 1}, \"formula_version\": \"ai-only-v1\", \"financial_bounds\": {\"revenue_min_jpy\": 400000.0, \"revenue_max_jpy\": 500000.0, \"profit_min_usd\": null, \"profit_max_usd\": null, \"profit_min_jpy\": null, \"profit_max_jpy\": null}}", "provenance_json": "{\"provider\": \"manual\", \"model\": \"beam-free\", \"confidence\": 0.5, \"evidence\": {}}", "source_hash": "6a515e161e7279064e38e59aa7d0ac4d2dc608643e2298953d857890ef99e93b", "analyzed_at": "2026-10-31T03:55:00.496292+00:00", "analysis_version": "offline-v1"}